The minimum wage might be stuck at a paltry $7.25 per hour on the federal level in the US, but all the 50 individual states in the union don’t have to take that lying down. At least, half of them don’t: In 2024, 25 US states will be raising their minimum wages.

That’s a big impact in total, with freshly boosted wages ranging from Montana’s new $10.30 per hour wage to Washington state’s impressive $16.28 per hour minimum.

Are you among the millions of American workers who will benefit at least a little from higher wages in 2024? Check out the list below and you’ll find out.

States Raising Minimum Wages in January 2024

Workers are still fighting for an economy-boosting $15-per-hour federal wage, but workers in three US states will actually surpass $16 per hour when the new year rolls around: Minimums in California and New York are both hitting $16 in January 2024, with only Washington’s $16.28 per hour standard representing a higher minimum.

Sadly, out of the 25 states that are not raising their wages in 2024, a full 20 haven’t budged from the federal minimum.

If your state isn’t on the list below, odds are not in your favor. The only place where our federal minimum wage is actually the “living wage” that it is meant to be was in the last decade in which it actually kept up with inflation: The 1960s.

In alphabetical order, these are all the US states that are raising their minimum wages on January 1, 2024, alongside that state’s new 2024 minimum wage.

• Alaska: $11.73 per hour
• Arizona: $14.35 per hour
• California: $16 per hour
• Colorado: $14.42 per hour
• Connecticut: $15.69 per hour
• Delaware: $13.25 per hour
• Hawaii: $14 per hour
• Illinois: $14 per hour
• Maine: $14.15 per hour
• Maryland: $15 per hour
• Michigan: $10.33 per hour
• Minnesota: $10.85 per hour
• Missouri: $12.30 per hour
• Montana: $10.30 per hour
• Nebraska: $12 per hour
• New Jersey: $15.13 per hour
• New York: $16 per hour
• Ohio: $10.45 per hour
• Rhode Island: $14 per hour
• South Dakota: $11.20 per hour
• Vermont: $13.67 per hour
• Washington: $16.28 per hour

Check your city’s wage minimums, as well: In many cases, workers within a major city’s limits will have a higher minimum wage in order to address the higher cost of living in a major city center.

States Raising Minimum Wages Later in 2024

Not every state’s new minimum standard for wages will go into effect on the first day of the year: Three of them are set to kick in a little while down the road in 2024.

So, if you’re in one of these three states, you’ll have to wait a while longer to get your higher wages.

• Florida: $13 per hour, as of September 30, 2024. Until then, it’s $12 per hour.
• Nevada: $12 per hour, as of July 1, 2024. Until then, it’s $10.25 per hour (if the employee offers qualifying health benefits).
• Oregon: TBD, as of July 1, 2024. Until then, it’s $14.20 per hour.

Those living and working in Orgeon state will be getting a new minimum wage on July 1, the same time they get it every year. Last time, on July 1, 2023, that adjustment was $14.20, up from the $13.50 minimum that was set on July 1, 2022. It’s based on the increase to the US City Average Consumer Price Index for All Urban Consumers, so keep your eyes on that index for more info.

Don’t Forget to Check Your State’s Overtime Regulations, Too

The US takes a famously hodgepodge approach to wage regulation, with every state settling on different minimums, plus a heaping handful of confusing add-ons. Overtime rules are another big factor worth Googling for your personal state, since there are so many variations.

For instance, if you’re in Nevada and make more than one and one-half times the minimum wage, you should be paid overtime at one and one-half times your regular rate of pay for every hour worked past your first 40 in any given week.

Plus, some states have different minimums for white collar workers as well.

Details of Sony’s video game roadmap, budgets, and corporate strategy are among the 1.3 million files leaked by cyber criminals in the latest data breach to plague the Japanese technology giant and, in this case, its Insomniac Games unit.

The breach was first announced by the Rhysida ransomware gang on December 12, at which time they said they would auction the huge cache of over 1TB worth of Sony files for $2 million paid in Bitcoin.

Now, the group has dumped the entire haul online – including details of a number of its upcoming Marvel titles from Insomniac Game – after failing to strong arm Sony or entice a third-party to buy the data. Here are all the details of the Insomniac Games leak and latest 2023 Sony attack in general.

Insomniac Games Leak Gives Gamers a Crystal Ball

For many people, specifically gamers, the headline of this December’s Sony data breach is the leaking of juicy details regarding its upcoming video game roadmap.

Sony’s Insomniac Games, which is under the company’s PlayStation Studios arm, is responsible not only for some of the console’s biggest AAA games like Ratchet & Clank, but also boasts a licensing agreement with Marvel to develop titles inspired by its franchises.

As first reported by CyberDaily.au, this means that the Insomniac Games leak is a treasure trove of upcoming Marvel game details – specifically, the apparent release windows for its first X-Men inspired titles. Three games have been agreed with a budget of approximately $120m a pop, with the first one taking on the Wolverine story.

• Wolverine (“no later” than September 1, 2025)
• X-Men game two (“before” December 31, 2029)
• X-Men game three (“before” December 31, 2033)

That’s not all, though. In addition to these claimed dates and the presumably once confidential budget figures for the games, the Sony Insomniac Games leak gives us even more details about what’s next for some of the studio’s biggest titles.

It’s not just tiles and release dates that have been leaked. Concept art, footage, and even playable early builds of some of these titles have also been included in the leaked data, meaning that there’s very little the studio is working on that isn’t now already out there for all to see.

Sony’s Insomniac Games Breach Reveals New Spider-man, Ratchet & Clank Titles

The amount of information now floating around online regarding Sony’s future gaming roadmap is vast. Over 1TB, in turns out, is a whole lot of leakage.

The highlights will be obvious to many gamers, however, and as well as the above Marvel titles include a new Spider-man game on the horizon that will be a follow-up to 2023’s Spider-man 2 and presumably complete the trilogy.

This one hasn’t got a timeline attached to it, but what it does have is the tantalizing prospect of being able to control Venom in the game. It’s also suggested that this might end up being drawn out into a “Part I” and “Part II,” making for four Marvel Spidey games in total.

What does have a purported release date is the next Ratchet & Clank game, which is said to be pegged for 2029 and be a direct follow-up to the last title in the series, Ratchet & Clank Rift Apart.

Sony Breach Also Lays Corporate Fears Bare

While it’s difficult to deny that a Wolverine game and new Spider-man titles are what we really want to read about as we wind down for the holidays, the Rhysida ransomware dump has also given a fascinating insight into some of Sony’s corporate strategy and thinking.

Specifically, the data breach has leaked slides that show Sony’s response to Microsoft’s recently approved acquisition of Activision.

It’s scared, in a word, especially when it comes to the impact of Microsoft essentially buying the wildly popular Call of Duty franchise as part of the blockbuster deal. As part of this acquisition,  Microsoft will likely be adding major games to its GamePass subscription service on the day of release in the future, and that’s something that Sony acknowledges it will struggle to compete with.

In other words, having a steady pipeline of new Marvel games is great, it’s just maybe not as great as owning probably the world’s best loved franchise.

Apple’s “Family Sharing” plan, which lets multiple people use the same apps, is the target of a new lawsuit. If you’ve used the plan in recent years, you might be owed a (tiny) chunk of a $25 million payout.

Anyone who used Family Sharing between June 21, 2015 and Jan. 30, 2019 and purchased a non-Apple app within that timeframe is eligible for their part in the settlement, which will likely add up to about $30 per person.

This isn’t even the first lawsuit Apple has settled this month, but it is a sizeable payout: We covered a gift card-related settlement just last week that added up to $1.8 million in total. Now, there’s a pot of free money worth over ten times that much. Here’s how to get your share.

What’s This Apple Lawsuit About?

The class action lawsuit alleges that Apple didn’t play fair: The Family Sharing plan is advertised as allowing family members to use any apps they want, but the service allowed app developers to opt out if they wanted — ensuring that not every app was actually available to subscribers.

That was enough for a big class action suit, and enough for Apple to pay for the privilege of settling the whole matter out of court.

So, if you were one of the poor suckers who wasn’t able to share with your family every single app that you paid for the right to share, you can now be compensated for your loss. However, you’ll still need to jump through a few hoops.

How to Cash In On Apple’s $25 Million Settlement

Hoping to claim your entitlement? Read on to learn more about how to make a claim.

Am I eligible for Apple’s settlement?

There are just three criteria that you’ll need to meet in order to qualify for this settlement:

1. Have been enrolled in a Family Sharing group with at least one other person at any point between June 21, 2015 and January 30, 2019.
2. Were a U.S. resident during that time period
3. Purchased a subscription to an app (that wasn’t published by Apple) through the App Store, also during that time period.

If you meet all three of those measures, you are able to ask for part of the settlement — there’s no further stipulation on which apps you purchased.

How do I make a claim?

You can file your claim online. Just go to the Settlement Website and click the link within the sentence “You can choose to receive a Class Payment by clicking HERE.”

You’ll need to provide some information alongside the form, so make sure to have the following to hand:

What information do I need to provide?

First, the claim process will ask if you have a “Payment ID and PIN,” which will be available to everyone who has received a notice about the class action suit. You may need to check your mailbox for one. If you have the ID and PIN numbers, you can fill them out online.

You can still file without this data, however! You’ll just have to print out and mail a payment election form instead. Here’s what data you’ll need to enter:

• Full name
• Primary address, including city and state
• Zip code

What’s the deadline for Apple’s settlement?

The deadline to file a claim is March 1, 2024. Your claim will not be processed if filed after this date.

How much money could I receive?

The exact number isn’t clear yet, but estimates put the potential payout at around $30.

When could I receive compensation?

The final approving hearing is on April 2nd, 2024. If all goes well, you should receive your settlement sometime afterwards.

If you want to embrace flexibility without stunting your career in 2024, you’re in luck. Apple is currently offering dozens of high-paying remote positions for US workers, bucking the trend many companies are making by rolling out mandatory office returns.

Famed for its inclusive hiring efforts, commitment to innovation, and unique employee perks, working at Apple is a tour de force for most tech enthusiasts. Combine this with a relaxed work-from-home policy (for certain roles!), and you may well be entering dream job territory.

If you’re interested in securing a job at Apple – all from the comfort of your home office – we’ve found some desirable positions that you can apply for today.

Fully Remote Apple Job Vacancies

From part-time gigs for students to vacancies for experienced engineers, here are some fully remote jobs Apple is hiring for right now – including their salary range, desired qualifications, and designated region.

Student jobs

Apple offers great opportunities for students looking for a part-time money-maker. Here are few open roles currently available.

1. Apple Support College Program At Home Advisor

Apple is currently hiring for support college program at-home advisors in over 30 colleges including big names like the University of Texas and smaller institutions like Bennett College. This is a customer-facing role that requires technical expertise and a passion for consumer technology.

• Weekly hours: 20
• Pay: From $27.75 to $32.45 per hour
• Qualifications: Must be enrolled at a four-year accredited US university pursuing a bachelor’s degree or higher through December of 2025 or later Technical, business, and communications majors preferred 2.7 GPA preferred
• Availability: Must be available to work three, four-hour shifts during the week between the hours of 4 pm and 10:30 pm, and one eight-hour shift on the weekend between the hours of 7:00 am and 10:30 pm.

Apply for the role

Software and Services jobs

Apple is also hiring for fully-remote engineers. Here are the two positions currently open:

2. Apple Home/Matter Software Engineer

This vacancy is open for a hardworking engineer to help build Apple’s Home automation platform.

• Weekly hours: 40
• Pay: Between $131,500 and $243,300 per annum
• Main qualifications: You need to be an accomplished coder in C, and object-oriented languages, such as C++ Objective-C, or Swift, and have a firm grasp of computer science fundamentals.
• Education: Bachelor’s degree in Computer Engineering, Computer Science, or equivalent experience.
• Reporting to: Office in San Diego, California

Apple for the role

3. ISE, System Experience – Software Engineer, macOS System Experience

The successful candidate will be working with the macOS System Experience team, navigating a large codebase, driving features forward, and solving complex problems creatively. Read on to learn more about the opening.

• Weekly hours: 40
• Pay: Between $170,700 and $300,200 per annum
• Main qualifications: Experience with  Swift, SwiftUI, and Objective-C, forward-thinking collaborator, customer-focused
• Education: Bachelor of Computer Science or equivalent work experience
• Reporting to: Offices in Santa Clara Valley, California

Apply for the role

4. Vulnerability Researcher

This role focuses on securing Apple products – from microarchitecture to applications like Safari and Mail. One of your main responsibilities will be to conduct vulnerability research on Apple software, firmware, and hardware components.

• Weekly hours: 40
• Pay: Between $170,700 and $300,200 per annum
• Main qualifications: Experience in vulnerability research and binary exploitation, and understanding of common vulnerability classes and exploitation techniques.
• Education: Degree in Computer Science / Computer Electronics (or equivalent experience)
• Reporting to: Offices in Santa Clara Valley, California

Apply for the role

Sales and Business Development jobs

If you have experience in sales and advertising, take a look if any of these roles appeal to you.

5. Healthcare Enterprise Account Executive, Major Accounts

The successful applicant will be responsible for advising on transformational solutions with our partners in Healthcare for Apple’s Enterprise Major Accounts spanning across the Midwest, developing account plans aligned with customer strategies, and more.

• Weekly hours: 40
• Pay: Not specified
• Main qualifications: Proven consultative selling experience in business transformation solutions,  8+ years as an Account Executive for large Enterprises
• Education: Bachelor’s degree or equivalent experience
• Reporting to: Offices in Cleveland, Ohio

Apply for the role

6. US AirPods & Accessories Sales Lead

The right applicant will be responsible for managing a team of experienced Beats and Accessories Account executives while creating channel marketing activities to achieve financial goals. Think this sounds like a bit of you? Take a look at the job spec below.

• Weekly hours: 40
• Pay: Between $169,100 and $254,100 per annum
• Main qualifications: 5+ years of worldwide business development experience in driving mature brands, extensive experience across sales, marketing, finance, and business development in the consumer electronics industry
• Education: Bachelor’s degree or equivalent experience required, Master’s degree (MBA or JD) preferred
• Reporting to: Offices in Santa Clara Valley, California

Apply for the role

7. Carrier System Engineer

You’d be joining Apple’s Carrier SE team to tackle technical and business problems for the company, the carrier, and the carrier’s customers. Specifically, you’d be supporting Apple’s carrier partners B2B channel, customers, and HQ.

• Weekly hours: 40
• Pay: Between $131,300 and $197,400 per annum
• Main qualifications: Strong technical background (15+ years) in engineering. Experience with evaluating complex IT enterprise environments (Deployment, Security, SSO, Zero Trust) and Mobile Technology.
• Education: Bachelor’s degree or equivalent experience
• Reporting to: Offices in Santa Clara Valley, California

Apply for the role

Corporate Functions jobs

Apple also offers one remote role within its corporate function division. Read on to see if it’s a good match.

8. Apple Professional Learning Specialist

Apple is on the hunt for an experienced learning specialist to provide leadership and planning support, build professional learning plans, and offer hands-on instructional coaching and mentoring to support teachers.

• Weekly hours: 40
• Pay: Between $131,300 and $197,400 per annum
• Main qualifications:
• Proven experience as both a teacher and leader at the K-12 level; Instructional design and coaching experience in a Higher Education setting or one-to-one learning environment is a plus. Familiarity with New York’s state teaching, technology, and content standards.
• Education: Bachelor’s degree required, Master’s degree or Doctorate preferred
• Reporting to: Offices in New York City, New York

Apply for the role

Can’t Find a Remote Position That’s Right For You?

There’s no denying that Apple is a great company to work for, with plenty of perks. However, since the company’s notorious U-turn on remote work, there have been fewer opportunities to join the company fully remotely.

This doesn’t mean that other great opportunities don’t exist, though. If you’re committed to joining Silicon Valley’s finest, Microsoft is currently hiring for hundreds of remote roles, while Google is also expanding its global, remote workforce.

If you’re not interested in joining the likes of Apple and big tech, and would like to pursue a healthier work-life balance instead, less demanding alternatives also exist. Take a look at six low-stress remote jobs that are available right now, for a more balanced 2024.

Alternatively, why not make your current job a remote one? Read our guide on how to ask your manager to work from home.

Gaming publisher Activision Blizzard has agreed to pay out $54 million to settle a 2021 gender discrimination lawsuit involving allegations of sexual harassment and pay disparity within the company, which is now under Microsoft after a blockbuster acquisition earlier in the year.

The settlement represents Activision Blizzard’s latest legal controversy, with previous lawsuits striking out against the company for its market monopoly and internal “frat boy” culture.

The company will be sharing $47 million of this sum with claimants who were unfairly paid. Read on for details about Activision Blizzard’s latest case, its eligibility criteria, and whether you’re able to file a claim.

Activision Blizzard Settles History Sexual Discrimination Lawsuit

Responsible for hit games like Call of Duty and Candy Crush, Activision Blizzard has decided to pay out $54 million to settle a gender discrimination lawsuit.

The suit, which was first brought against the company in 2021 by the California Civil Rights Department, revolves around allegations of sexual harassment, denied promotion opportunities for female workers, and pay disparities among genders.

If approved, this will be the second-largest settlement proposed by the California agency, behind Riot Games’ $100 million class-action payout in 2021, which involved similar claims of gender-based discrimination against female employees.

The majority ($47 million) of Activation’s payout will go to female employees and contractors who were treated unfairly from October 2015 to December 2020, while $9.1 million will be reserved to cover legal fees.

The landmark case hasn’t been without its difficulties, though. Since the class-action lawsuit was brought against Activation Blizzard, the company has repeatedly denied charges and disregarded claims as false. The final settlement sum also dwarfs the state agencies’ initial liability estimate of $1 billion.

Microsoft Cashes in on Activision’s Controversy

High-profile allegations of sexual discrimination, alongside news that company CEO Bobby Kotick purposefully hid accusations of sexual misconduct from the board, caused Activision’s stock price to tumble in 2021.

The company’s subsequent fall in evaluation provided a golden opportunity for Microsoft, however, with the software manufacturer announcing a takeover bid in January 2022.

After fending off global competition, Microsoft officially acquired Activision for $69 billion in October this year, securing it’s position of the third-largest gaming company in the world, after Tencent and SONY.

Who is Eligible to Claim in the 2023 Activision Blizzard Settlement?

If you’re a female worker, who worked at Activision Blizzard California between October 12, 2015, and December 31, 2020, you may be eligible to receive a slice of the company’s $54 million settlement.

However, since the case hasn’t yet been officially approved, no action needs to be taken by affected employees. Eligible workers will receive further information from a settlement administrator, and regular updates will be posted on the California Civil Rights Department website

We’ll update this article with additional information as it becomes available.

As artificial intelligence continues to streamline business processes, AI transcription services are being used more than ever to transcribe audio and video recordings into written text.

Not only do they generate results faster than manual transcription — which can often be a tedious, time-intensive process — they’re also more affordable, with loads of low-cost and free services available to desktop and smartphone users.

If you’re interested in capitalizing on artificial intelligence without spending a dime, this guide covers quality free AI transcription tools, including their key features and device compatibility. Scroll down to discover a ranking of our top picks or skip to a specific transcriber using the links below.

The 7 Best Free AI Transcription Tools

Whether you’re a podcaster or the designated meeting scribe, here’s our round-up of the seven of the best free AI transcription tools:

1. Otter AI
2. Riverside
3. Happy Scribe
4. Vowel
5. Descript
6. Grain
7. Notta

1. Otter AI

• Best for: Students and workers transcribing notes
• Compatibility: Mac, Windows, iOS & Android

Otter AI is an all-in-one AI meeting assistant that’s been designed with the needs of note-takers in mind. Its generous free plan lets users record, transcribe, and summarize meetings and lectures in real-time and also boasts integrations with Zoom, Microsoft Teams, and Google Meet.

Otter’s free transcription cap is much more generous than other services, allowing for 300 monthly transcription minutes each month (at 30 minutes per conversation), from three audio and video files.

Otters AI Chat feature is available on its free plan too, allowing users to chat to other teammates instantaneously. If you need to transcribe video or audio files regularly, however, you’ll have to bump up to Otter’s Pro plan at $10 per user, per month, which lets you import and transcribe 10 files per month.

2. Riverside

• Best for: Podcasters and journalists transcribing interviews
• Compatibility: Mac & Windows

Riverside is a free recording platform that offers audio and video-to-text transcriptions in over 100 languages. It claims to transcribe content with 99% accuracy, which is a huge boon for a free tool, and also offers recording and editing features at no extra cost.

Riverside AI transcription screenshot. Source: riverside.com

These extra features make Riverside ideal for content creators looking for an all-in-one recording, editing, and transcribing tool. Yet, its speaker differentiation feature, which automatically displays when different people are talking, also makes the tool useful for workers transcribing complex conference calls involving multiple speakers.

However, Riverside’s free plan only allows for two hours of audio and video transcriptions, on a one-off basis. What’s more, in order to download transcripts, you’ll have to upgrade to one of its paid plans — making the service less suitable for users transcribing large amounts of content.

3. Happy Scribe

• Best for: Beginners to transcribing tools
• Compatibility: Mac & Windows

Happy Scribe is a user-friendly transcription tool that offers both automatic and human subscriptions. Its free AI transcription service is available in over 120 languages and can deliver results in under five minutes.

Happy Scribe transcription service. Source: happyscribe.com

The tool accepts a wide range of audio and video input, including desktop files, YouTube URLs, and Google Drive links, and also lets users share transcripts in different formats, with the option to activate view-only or edit modes.

Its free plan offers in-house subtitle and translation services too, but only 10 minutes of your recordings are free. Its AI-powered service only claims to be 85% accurate as well, trailing other services like Riverside and Notta.

If you’re transcribing longer files, and want greater accuracy, you’ll have to upgrade to its paid plans with human-made transcription services and be willing to wait 24 hours for results.

4. Vowel

• Best for: Managers and video call hosts 
• Compatibility: Mac, Windows, Google Play & App Store

Vowel is an AI meeting helper that lets you host, record, and transcribe meetings. Vowel is one of the only platforms on this list that offers hosting capabilities in-house, making it one of the best choices for managers and other workers tasked with logging minutes in meetings.

Vowel meeting transcriber tool. Source: vowel.com

Vowel works in a similar way to web conferencing services like Zoom but goes above and beyond by offering AI-powered meeting summaries, and advanced live transcription services.

This makes the tool one of the best options out there for users looking to transcribe video calls. However, users looking to transcribe other types of content like interviews or video and audio files will be better off using Otter or Riverside instead.

5. Descript

• Best for: Podcasting and content creator beginners 
• Compatibility: Mac and Windows

Descript is an app for podcasters and content creators that lets you record, edit, and transcribe video and audio content. Descript offers a number of unique features like filler word removal and studio sound quality.

Descript editor and transcriber. Source: descript.com

Descript also has an AI-powered overdub feature that lets you edit text to create voiceovers in your own voice, offering a huge advantage for users who aren’t comfortable speaking themselves.

However, if you’re just looking for a straightforward transcription tool, Descript’s input cap of one hour a month may put you off. If you want a more generous time limit, we recommend upgrading to its Pro plan, at $24 per user, per month, or using other services like Otter AI.

6. Grain

• Best for: Transcribing and summarizing work meetings
• Compatibility: Windows, Mac & Google Play

Grain is an AI conversation intelligence assistant that lets users record, transcribe, and highlight key information in video calls.

Grain’s free plan, Basic, has a much more generous transcribe cap than services like Notta, letting users transcribe 20 meetings, and five additional uploads a month. Free users have access to Gain’s AI notes feature too, which summarizes key parts of the meeting.

Grain AI transcribing and notes feature. Source: grain.com

Grain only lets you transcribe meetings in nine different languages, compared to other services like Riverside and Happy Scribe, which offer over a hundred different language options. This makes the platform less suitable for businesses with a global reach and prevents it from appearing any higher on our list.

7. Notta

• Best for: Workers recording short meeting snippets
• Compatibility: Windows, Mac, Google Play & App Store

Notta is a free transcribing service that lets you convert video and audio content into written transcripts in minutes. It claims to transcribe content with an impressive 98.9% accuracy rate, giving leading services like Riverside a run for their money.

Like our frontrunner Otter, Notta also integrates with leading services like Zoom, Google Meet, and Microsoft Teams, making it easy for workers to record and transcribe work calls.

Notta also offers lots of advanced free features, like screen recording, speaker identification, and live recording transcription.  The free service only lets users transcribe content up to three to five minutes at a go through, which will be a dealbreaker to users wanting to transcribe entire meetings.

Should I Use a Free Transcription Service?

Thanks to these free tools, businesses and personal users no longer have to pay a lot for high-quality, speedy transcription services. However, while these platforms offer impressive accuracy and a boatload of other AI-powered features, their stingy user caps make them ill-suited to users looking to transcribe high volumes of content.

With this in mind, those with regular transcription needs will be better off paying for paid plans. Premium plans tend to offer better quality transcription services too, and often end up being well worth the extra dollars per month.

OpenAI’s “Superalignment” team has announced a small breakthrough in how humans could be able to reign in AI once it supersedes our level of intelligence – a possible scenario known as AGI superintelligence.

Researching superintelligent machines has limitations though, as even once the technology becomes realized it may try and hide its true behavior from humans, according to Ilya Sutskever – OpenAI’s Co-founder and the driving force behind Sam Altman’s temporary ousting as CEO.

While superintelligence still remains hypothetical, OpenAI appears to be taking the concerns seriously, dedicating a 5th of its computing power to risk mitigation and investing $10 million into super alignment research.

OpenAI’s AGI Test is Promising, But Has Flaws

As OpenAI becomes increasingly concerned about the looming threat of AI superintelligence, its Superalignment team has just released its first research update – and its results are mixed.

The team, which was founded by Ilya Sutskever and OpenAI scientist Jan Leike, looked into how superintelligence systems could be supervised once they surpassed human capabilities, but since these machines don’t exist yet, OpenAI’s control test used GPT-2 and GPT-4 systems as stand-ins.

More specifically, researchers tested how the less sophisticated GPT-2 model would be able to supervise its most powerful model GPT-4 model. But was was their conclusion?

Well, after training GPT-2 to perform different tasks, including chess puzzles and sentiment analysis, and using these responses to train GPT-4, they found that OpenAI’s latest model performed 20-70% better than GPT-2, but still fell far short of its own potential.

GPT-4 also avoided many mistakes made by the inferior GPT-2 model. According to the researchers, this is evidence of a phenomenon called ‘weak-to-strong generalization’ that occurs when a model has implicit knowledge of how to perform tasks, and is able to execute them correctly despite receiving poor-quality instructions.

According to researchers, evidence of weak-to-stong generalization in the test suggests this phenomenon may also be present when humans supervise superintelligent AI. If this is true, when AGI models exist, and are instructed not to cause catastrophic harm, they may be able to tell even better than their human supervisors when their actions enter dangerous territory.

However, despite these positive findings, GPT-4’s performance was still weakened after being trained by GPT-2, and the experiment didn’t guarantee the stronger model would behave perfectly under supervision. These shortcomings reveal that more research has to be conducted on the matter before humans can be trusted as suitable supervisors.

OpenAI: Concerns Around Superintelligence Are “Obvious”

Despite being a frontrunner in AI development, OpenAI hasn’t been shy in voicing the potential dangers of superintelligent AI.

Co-founder Sutskever recently came out saying preventing AI from going rogue is an “obvious” concern, and the company even announced in a statement that technology could be very dangerous, and could lead to the disempowerment of humanity or even human extinction if left unchecked.

“We’re gonna see superhuman models, they’re gonna have vast capabilities and they could be very, very dangerous, and we don’t yet have the methods to control them.” – Leopold Aschenbrenner, OpenAI Researcher

But why are researchers so scared of these machines going rogue? Well, according to Sutskever, once superhuman AI models exceed human-level intelligence, they could become capable of masking their own behavior, opening us up to worrying and unpredictable and worrying circumstances.

OpenAI Claim to Be Taking AI Safety Seriously

While many experts believe these fears are overblown, OpenAI is already taking several steps to address safety concerns.

The company recently announced it would be investing $10 million into super alignment research, in the form of $2 million grants to university labs, and $150,000 grants to individual graduate students. Open AI also revealed it would be dedicating a fifth of its computing power to the Superalignment project, as it continues to preemptively research how to govern AGI.

Disputes around superintelligence and AI safety were even rumored to be the real reason behind Sam Altman’s shock, yet brief, deposition from the company last month. However, with  Altman currently back at the helm, and OpenAI continuing to prioritize commercial interests with its AGI project ‘Q*’, the company could still be doing a lot more to to mitigate risks to public safety.

How can you tell remote jobs are here to stay? Because even after spending all of 2023 pushing back against the concept, the biggest tech companies in the business still have hundreds of 100% work from home positions available.

Microsoft is one of the biggest and oldest names in tech, with a rich history that includes that one time Bill Gates jumped over a chair. Probably some other stuff, too. Look, the point is that the Microsoft name is a great resume builder. Joining up connects you to a high-quality network of tech professionals at every level.

And, with around 890 jobs currently available for remote workers, it’s easier than ever to land yourself a work-from-home position at the Redmond, Washington-based corporation, all without leaving your living room.

Fully Remote Microsoft Jobs: Listings and Locations

Settling into a remote job at Microsoft will be a lot easier if you have a lot of subject authority as a senior engineer or top-level management professional.

But you won’t have to be: Available roles include titles like a Customer Success Account Manager based in San José, Costa Rica (you can still apply even if you’re nowhere near the country) or a Regional Delivery Manager across multiple locations in the US.

Here’s a quick example list of some of the fully remote jobs available at Microsoft:

• Customer Success Account Management (San José, Costa Rica)
• Senior Software Engineer – Azure Storage (Redmond, Washington, US)
• Principal Mechanical Architect (Multiple Locations, US)
• Advertising Account Executive, Travel Industry (Seattle, Washington, US)
• Customer Experience Engineering Manager (Redmond, Washington, US)
• Fiber Delivery Engineer (Multiple Locations, US)
• Director Planning Infrastructure – Systems Applications & Products (Redmond, Washington, US)
• Technical Program Manager (Multiple Locations, US)
• Cybersecurity Incident Response Engineer, Manager (Multiple Locations, UK)
• Senior Content Designer (Redmond, Washington, US)
• Software Engineer – Sales Copilot (Redmond, Washington, US)
• Regional Delivery Manager (Multiple Locations, US)

Microsoft currently has just 408 “job site only” positions available, compared to 893 openings for “up to 100%” work from home positions and a huge 1,726 available jobs for the hybrid-working “up to 50% work from home” crowd.

That’s a huge sign that Microsoft is remote-friendly. After all, we just reported on remote positions at Google that are available this month — and found just 60 fully remote openings at the search giant out of a total of 1,000 open jobs. 

Is Remote Work Right for You?

Remote work is not for everyone. Some people try it out only to find they hate the lack of a structured routine that a well-worn commute or a trip to the office break room can provide. And you’ll need to be self-motivated in order to keep plugging away at a nine-to-five without direct supervision outside of a Slack channel.

However, many people find they work far better when the trip to work takes two minutes, and the hubbub of an office building isn’t distracting them.

Plenty of remote work statistics support the concept, too: Employees report more work-related optimism, while 47% of businesses report boosted productivity among their work-from-home employees.

You’ll want to keep an eye out while job-searching, however: Some positions require in-person travel, while others are unclear on how much remote work will be expected alongside in-person office appearances.

These concessions to physical office space might be reasonable for most workers, but not everyone has even a little flexibility to visit an office for one or two days a week, or even to take a company retreat once or twice a year. The number of Americans with disabilities is higher today than it has ever been since the government started tracking the data in 2008. The Covid-19 pandemic is likely a big cause, given the hike in numbers since it started: The U.S. Department of Labor found in 2022 that the total of disabled adults in the nation had risen by 2.7 million since 2020.

For those with serious reasons to stay entirely remote — from childcare needs to physical disabilities — even one day a week in a physical office can derail everything.

Should You Work at Microsoft?

One thing’s for sure: You’ll need to know Microsoft Teams inside and out. As the company’s personally branded method of syncing up in-person and remote workers alike, Teams is going to be your main tool for tasks like internal communication and meeting schedules. Microsoft has even used Teams to run a massive 500-person live global meeting that spanned an impressive 48 different countries.

But is Microsoft truly remote-friendly, or is it all a PR stunt? Commenters on the Microsoft subreddit seem to be largely affirming that the company really is happy to employ entirely at-home workers, even if they’re in another state or country outside of Washington state in the US. In fact, the company-wide accessment of remote work is that nearly anyone can do it:

“The official policy is that everyone can work from home up to 50%, or 100% if you have management approval (some roles aren’t eligible to WFH, but they’re rare)” –Reddit commenter

A few caveats worth keeping in mind: Microsoft will adjust your wages to stay competitive with the area you live in, rather than the area you’re remotely commuting to. Within the US, there isn’t a huge difference, but if you’re in, say, the UK, Microsoft may offer you a lower salary than they’d offer an equivalent US-based role.

On the whole, though, Microsoft doesn’t seem likely to pull the rug out from under you by revoking their remote-friendly policies — even if steady remote work is getting a lot tougher these days.

How to Apply for Fully Remote Jobs at Microsoft

If Microsoft sounds like a fit for you, you’re ready for the next step: Filling out a few (or a few dozen) online applications.

You can visit Microsoft’s dedicated job opening website now and search for open positions by a variety of metrics including profession, discipline, and employement type. On the “work site” dropdown, you’ll see options for “Up to 100% work from home,” “Up to 50% work from home,” and “Microsoft job site only,” letting you chose how much remote work fits your work style.

Midjourney, one of the most popular image-generating AI services, announced yesterday the launch of the “Alpha” version of its website, available to a select few users.

The San Francisco–based independent research lab announced they will be testing out the site, Midjourney.com. It is currently only available to users who have generated over 10,000 images on its predecessor, messaging app Discord.

The tech company confirmed that the website will roll out to the rest of us soon, the announcement read: “Don’t worry, we will be making it available to more people over the coming month!” This will be music to the ears of the 17.5 million plus users in its server on the Discord App.

In this guide you’ll learn about Midjourney, including the new and improved features available and how you can use it to start creating images.

What is Midjourney?

Midjourney is a generative artificial intelligence program which understands user prompts to create incredibly lifelike and realistic images.

The technology uses natural language descriptions, known as prompts, similar to OpenAI’s DALL-E and Stability AI’s Stable Diffusion.

How Much Does Midjourney Cost?

Tragically, the service is no longer free since April 2023. The free trial had to be removed as the popularity of the tool led to overloaded servers. However, limited time free trial periods may pop up from time to time, so keep an eye out if you’re looking to try before you buy.

There are different pricing plans available, ranging from $10 per month for the most basic and fairly limited plan, up to $120 per month for the Mega plan with all bells and whistles. Discounts are applied when a yearly subscription is taken out instead of monthly rolling payments.

New and Improved Functionality

The website has a modern interface and launches with an “Explore” tab, which demonstrates real-life user generations. At the top, you’ll find a plus sign and next to it is a familiar blank text bar that’s labelled “Imagine” just waiting for users to enter prompts.

On the right-hand side, users can click on the “dials” symbol, which will launch the visual interface where they control their prompt settings.

This is a major upgrade compared to the Discord version of Midjourney, where users had to type text commands into their prompts in order to modify them. Now users can quickly change the orientation, size, and stylization of the image easily.

On the left-hand rail, users can click on the “photo” icon to sort their past creations into a reverse-chronological order, as well as see their ongoing projects.

There is a “thumbs up” button for users to vote between two iterations of the same image in a crowdsourced image ranking project. Whereas the “Likes” tab contains all of the images users have previously “hearted” on Discord.

More features are reported to be coming out soon, including an app for iOS, folders, and collections.

How to Use Midjourney

Now that you know all about how Midjourney works and its new features and functionality, it’s time to start using it. But how can it be useful to you?

AI-image generation has a variety of use cases in the professional world, and in an increasingly visual world, it’s a welcome addition in the toolkit of almost any business.

Small businesses without design capabilities will be able to create promotional content and engaging social media posts at scale.

Content marketers will also save time and money on image sourcing. Rather than entering into the realm of creative commons, purchasing costly stock images, or using photographers and graphic designers, they will be able to create imagery bespoke to their needs.

Creative teams or graphic designers can use the tool to get inspiration and speed up their existing design workflow.

Some Dropbox customers got a shock this week when they found out a new setting had been enabled by default that allowed their personal data to be shared with ChatGPT owner OpenAI. If that’s not something you like the sound of, we’re here to show you how to stop Dropbox sharing your data with OpenAI.

Many disgruntled users took to social media to air their concerns about the feature, which shares Dropbox data (including documents and files) with OpenAI as part of an AI Alpha search tool. Dropbox stated that data shared with third-party AI partners isn’t used to train AI models and is deleted within 30 days.

Dropbox CEO Drew Houston took to X (formerly Twitter) to apologize recently and quell any fears over customer data privacy.

He wrote: “The third-party AI toggle in the settings menu enables or disables access to DBX AI features and functionality. Neither this nor any other setting automatically or passively sends any Dropbox customer data to a third-party AI service.”

How to Stop Dropbox Sharing Your Files and Data

Despite its AI and privacy FAQ page detailing the safeguards it has in place, many Dropbox customers still want to disable Dropbox sharing with OpenAI.

Fortunately, it’s easy to opt-out of Dropbox’s AI features, including data sharing with OpenAI. Just follow these steps:

1. Log into your Dropbox account and click on your profile photo
2. Navigate to Settings and then to Third-party AI
3. Click the switch next to “Use artificial intelligence (AI) from third-party partners so you can work faster in Dropbox” to switch the toggle “Off”.
4. Or use this link to access it quickly.

Should You Opt-in to Dropbox’s AI Features?

A Dropbox representative told Ars Technica, “The third-party AI toggle is only turned on to give all eligible customers the opportunity to view our new AI features and functionality, like Dropbox AI. It does not enable customers to use these features without notice. Any features that use third-party AI offer disclosure of third-party use, and link to settings that they can manage.

The statement continues: “Only after a customer sees the third-party AI transparency banner and chooses to proceed with asking a question about a file, will that file be sent to a third-party to generate answers. Our customers are still in control of when and how they use these features.”

Many users wouldn’t be opposed to the AI-powered features, especially those willing to embrace the new technology for its time saving superpowers. The incident has highlighted the importance of clear communication and transparency regarding AI and data privacy concerns for businesses.

What Do Dropbox’s New AI Features Actually Do?

So what does the feature do exactly? Dropbox unveiled new AI tools like Dash in July 2023, at the time explaining in a blog post how the universal search tool works.

Users can perform searches across their work platforms such as Google Workspace and Microsoft Outlook, using AI to connect the tools and save workers much-valued time. Dropbox cited a McKinsey report that found that knowledge workers spend 8.8 hours a week of unproductive time switching between apps searching for files and content.

Users are able to converse with the tool, much like a large language model (LLM) chatbot, in this case ChatGPT, asking it questions about the files.

The popularity of AI chatbots has exploded over the last year, with every big tech firm and small startup putting out an iteration that can answer questions, generate content, and perform a wide range of other impressive tasks.

However, with so many options out there, how do you choose which AI chatbat is the right one for you? Well, with Poe, you don’t have to! This AI chatbot aggregator allows you to access a wide range of AI chatbots in a single hub and even create some of your own.

In this guide, you’ll learn exactly what Poe is, how it works, which chatbots are available, how much it costs, and how you can make money on Poe by creating your own chatbots.

What Is Poe?

Poe is an online platform and app that allows you to access a wide range AI chatbots, including ChatGPT, Claude, and even is own, to more effectively aggregate your options into a single hub. Poe also has its own AI chatbot called Assistant, which you can add to your AI chatbot lineup on the platform.

On top of that, you can create your own AI chatbots to store on the platform, which can also be used by other users. You can even get paid for your chatbots when other users take advantage of your creations.

Which AI Chatbots Are Available on Poe?

What sets Poe apart from other AI chatbots is that it doesn’t just allow you to access its own chatbot, Assistant. Instead, you’ll be able to access, store, and organize conversations with other AI chatbots, creating a full-on hub for your conversations and content.

So, which AI chatbots are available on Poe? Here’s a comprehensive list of some of the most notable AI chatbots you can find on the platform:

• ChatGPT from OpenAI
• DALL-E-3 from OpenAI
• Google PaLM
• Llama 2 from Meta
• Claude Instant from Anthropic
• Claude 2 from Anthropic
• StableDiffusionXL from Stability AI

Poe has been frequently updating its platform to integrate with more official AI chatbots, so be sure to check back to see if your favorite gets added.

This list of official AI chatbots is just the tip of the iceberg for Poe, though. The platform also allows users to create their own chatbots and makes them available to all Poe users. These chatbots are specialized for more specific purposes, from creating stock photography to predicting sports scores.

Where Is Poe Available?

Poe is available across all the expect avenues, so you shouldn’t have any trouble with using the AI chatbot aggregator on your device.

For starters, Poe is available on web. Just head on over to poe.com and you’ll be able to get started once you login. There’s also a desktop app that, according to Poe, will give you faster functionality for AI chatbots.

Additionally, Poe comes as a mobile app for both iOS and Android, so you can aggregate all your AI chatbots on your smartphone for easy use.

How to Use Poe

Now that you know about everything that Poe has to offer and where you can use the service, it’s time to get started. Fortunately, Poe is extremely intuitive, with a simple interface that will have you engaging with AI chatbots faster than you can say, “Skynet.”

To kick things off, you’ll have to select which AI chatbot you actually want to talk to. You can access a list of your options in a few different ways, including clicking the Your bots button, the Explore banner, or the start chat icon found in the interface. This will bring up a list of AI chatbots that you can choose from.

Once you’ve decided which AI chatbot you want to talk to, you’ll be taken to a chat box that will allow you to ask questions and posit requests. The interface is the same for all chatbots, although you’ll obviously get different responses depending on which one you’ve selected.

If you want to experiment with a bunch of different AI chatbots, you can easily revisit former chats. There’s a menu bar that keeps track of all your different conversations, so you can easily swap between them to compare your answers.

How Much Does Poe Cost?

Good news! Poe offers a free version that will allow you to access a wide range of AI chatbots that can help handle tasks from image generation to content creation.

However, the free version is a bit limited, particularly when it comes to which AI chatbots you have access to. The newest versions of GPT-4, DALL-E-3, and Claude 2 are all behind a paywall, so you won’t be able to use the more advanced versions without subscribing with a monthly fee.

Poe comes with a single paid plan, which costs $19.99 per month, or you can sign up for a full year, which will bring the cost down to only $16.67 per month, which comes out to around $200 per year.

In addition to unlocking the AI chatbots listed above, the paid plan will get you faster responses from AI chatbots and will give you higher allotments of messages that you can send daily.

How to Create Your Own Bot on Poe

Poe is more than just a simple AI chatbot aggregator for official options from big tech firms. You can also create your own bots on the platform that will be available to users across the Poe community.

As for how to create bots, the process again is shockingly simple. Just click the Create a bot banner and you’ll be taken to a form that will get you started. You can edit the image, change the name, develop the prompts, select the base bot, and add the knowledge source, along with a few other customizable features for your bot.

Once the bot is ready, select the Create bot button, and you’ll be taken to a chat box like any other official AI chatbot, so you can ask questions and get responses based on your input. Then, you can share it across the platform, on social media, and with your friends with a chance to even make a bit of money if it becomes popular.

How to Get Paid from Poe

Fortunately, Poe doesn’t expect you to create your own AI chatbots without a bit of compensation. There are a few ways to get paid from Poe for creating AI chatbots, which can help you make a bit of money from your AI prowess. Here are the ways in which you can make money from Poe:

• When your bot brings users to Poe for the first time and they subscribe.
• When your bot brings users back to Poe and they subscribe.
• When your bot’s paywall is seen just before users subscribe.
• When users send messages to your bot (in development).

As we’ve mentioned before, developers are constantly updating Poe to make it more effective, which means that compensation packages will likely improve over time, so be sure to check back for updates in regard to how to get paid on Poe.

Etsy, the online marketplace which connects craft-makers and craft-lovers from around the globe is the latest in a slew of lay offs which have happened over the past year.

Etsy announced an 11% job cut yesterday, representing an attempt to streamline costs in what it has called a “very challenging” environment.

The e-commerce giant looks poised to undergo an overhaul of its structure as chief executive Josh Silverman seems laser focused on ensuring the future growth of the company.

Unsustainable Trajectory for Etsy

In the letter to his staff, CEO Josh Silverman blamed “essentially flat” merchandise sales for the past two years for the jobs cut.

His letter continued, “This means we are not bringing our sellers more sales, which is the single most important thing we can do for them. At the same time, employee expenses have grown, even as we have introduced significant cost-cutting measures and adjusted or paused hiring plans. This is ultimately not a sustainable trajectory and we must change it.”

“Increased competition” was among the factors cited by Silverman which contributed to the challenging macro environment. He did not reference competitors by name, however the likes of Shein and Temu, who offer consumers cheap manufactured goods from China and have recently flooded the market with aggressive online advertising strategies, come top of mind.

The news broke at the same time as an updated fourth-quarter guidance was published for Etsy. In the guidance report the company forecast that gross merchandise sales will decline between 1% and 2% during the fourth quarter of this year compared to the same period last year. Revenue figures look set to increase by 2% to 3%.

What’s more, shares in Etsy declined by a further 2.2% following the announcement from Silverman, falling as low as $78.54.

Poor Timing by Silverman

The news coming just two weeks before Christmas was unlikely to be a popular decision, however Silverman reassured affected workers that they will remain on Etsy payroll until “at least” January 2nd although many of them had their last day on Wednesday.

The cut has affected workers on all levels, with several executives including Ryan Scott, chief marketing officer, already announcing his departure.

The lay offs come at a cost of between $25m and $30m to the company in severance packages and other related costs, according to their regulatory filings.

Speaking about the announcement of Wednesday, Silverman said it was “one of the hardest [days] we’ve experienced at Etsy”, he continues addressing workers who will remain employed at Etsy.
“The waters may be rough right now, but there’s no other ship I’d rather be on, and no other crew I’d rather be with, as we weather this cycle and emerge even stronger on the other side.”

Want to know how to go about the thankless task of laying off staff? Here’s some advice for dealing with lay offs.

Internet speeds are getting out of hand, with Google announcing that its GFiber service will now offer a super-fast 20Gbps connection with Wi-Fi 7 — but it’s only available in a few areas.

Technology continues to evolve at break-neck speeds, and it’s no secret that more is on the horizon with the rise of AI chatbots in the world. As a result, faster and faster internet connections are required to power these platforms, devices, and other tech-driven innovations.

Fortunately, Google seems more than up to the task, announcing an insanely fast service that can handle more than you could possibly throw at it and more.

What Is GFiber’s 20 Gig + Wi-Fi 7 Service?

GFiber is a service from Google that provides internet connectivity to select areas of the country. It comes in a variety of pricing plans, which offer varying speeds depending on your particular needs.

Announced this week, Google is launching a new, super-fast GFiber plan that gives you 20Gbps speeds across your network. For some context, the fastest speed provided by GFiber before this was only 8Gbps, which means this new plan is the definition of speed.

“It’s fast across the board and enables multiple devices to take advantage of multi-gig speeds, even over Wi-Fi and across multiple floors — which is important because the vast majority of our in-home internet connections happen over Wi-Fi.” – Nick Saporito, Head of Product at GFiber in a Google blog post

On top of that, the GFiber 20Gbps plan will offer compatibility with Wi-Fi 7, a yet-tested standard that operates faster than previous iterations. Google worked with Actiontec to build out a router that can handle all this speed, even if it is pre-certified.

How Much Does GFiber Cost?

As you can imagine, these kind of internet speeds are anything but cheap. The new 20Gbps plan from GFiber will cost you $250 per month, which is quite a hefty sum to ensure that your speeds are the fastest on the block.

If you think you can get by on slower speeds, GFiber does have a range of pricing plans that might suit you better. The 1Gpbs plan will cost you only $70 per month, the 2Gbps plan costs $100 per month, the 5GB plan costs $125 per month, and the 8Gbps plan will cost you $150 per month.

Where Is GFiber Available?

It’s worth noting that GFiber is not available anywhere, even across the US. In fact, the high-speed plans, including the new 20Gbps plan, are available in limited areas across the country, so make sure you’re eligible before making any financial commitments.

The new 20Gbps plan is only available in Kansas City, Arizona, Iowa, and North Carolina (Triangle Region). There are obviously plans to roll out the technology further in the future, but no date has been set for any new regions.

Luckily, the 5Gbps and the 8Gbps are more widely available, although you’ll still have to check if your eligible before you buy. Some available cities include Huntsville (AL), Kansas City (MO), West Des Moines (IA), and Salt Lake City (UT).

If you don’t need the fastest Wi-Fi around, the 1Gbps and 2 Gbps plans are available in a lot more locations, so you can always go that direction.

Tesla is having to recall over two million cars in the US after the Department of Transportation’s National Highway Traffic Safety Association (NHTSA) found its Autopilot to be defective.

The company, owned by Elon Musk, launched its driver assistance system back in 2015, meaning the recall applies to almost every Tesla sold in the US since. 

This ruling follows a two-year investigation into crashes that took place when Autopilot was in use, and comes three months after the NHTSA sent a special order to Tesla demanding it hand over data relating to a risky secret car configuration known as ‘Elon Mode’.

Why Is There a Tesla Autopilot Recall?

According to Tesla, Autopilot is designed to “enhance safety and convenience” behind the wheel. It can help with actions like acceleration, steering, and braking but still requires input from the driver.

This is where the driving monitoring system comes in, as a safety feature to check things like whether the driver is paying attention and has their hands on the wheel.

According to the NHTSA, this recall is down to an issue with Autopilot’s monitoring system.

Earlier this week, Tesla defended the safety of its Autopilot in a post on X/Twitter. It stated that “safety metrics are emphatically stronger when Autopilot is engaged than when not engaged” and that “the data is clear: the more automation technology offered to support the driver, the safer the driver and other road users.”

After disagreeing with the NHTSA’s analysis, Tesla has since gone on to concede that the system “may not be sufficient to prevent driver misuse”. 

It went on to state that it would provide a software update that will “incorporate additional controls and alerts to those already existing on affected vehicles to further encourage the driver to adhere to their continuous driving responsibility whenever Autosteer is engaged.”

Which Tesla Models are Affected? How to Check if Your Tesla Has Been Recalled

As we’ve said, the 2023 Tesla recall applies to every model delivered since 2015 – or pretty much every Tesla on US roads right now.

However, to be more specific, Tesla has issued a recall notice for the following models:

• Tesla Model S
• Tesla Model X
• Tesla Model 3
• Tesla Model Y

If you’re still unsure if your vehicle is affected, you can check if your Tesla has been recalled by running your VIN through Tesla’s recall search.

Despite being referred to as a recall by regulators, drivers of affected cars will not need to visit a dealership or garage to have their software updated. Rather, Tesla has confirmed the update will take place ‘over-the-air’.

The update will include a number of changes to Autopilot, including amplifying alerts on the car’s user interface, simplifying Autosteer engagement/disengagement, and a new system whereby drivers will be blocked from using Autosteer is they fail to prove they are using it and driving responsibly.

Tesla Recall Background: Over 900 Tesla Autopilot Crashes Reported

This recall comes following a two-year investigation into Tesla collisions, sparked by its cars “[ploughing] into the rear” of a parked fire engine and parked police car on two separate occasions. It’s said that either Autopilot or a system called Traffic Aware Cruise Control were active just prior to the crashes.

These were only a couple of examples among a list of 11 cases that initially prompted the investigation. Following this, the NHTSA has since reviewed 956 collisions where Autopilot was alleged to have been in use.

Referencing the recent ruling, the NHTSA explained “Automated technology holds great promise for improving safety but only when it is deployed responsibly”. It also confirmed that it would continue to monitor Tesla’s software following this latest update.

Whistleblower Safety Concerns

This ruling follows a recent whistleblower report from the BBC in which a former Tesla employee spoke out about safety concerns around the vehicles. 

Lukasz Krupski had leaked data, including customer complaints relating to Tesla’s braking and self-driving software, after his attempts to highlight concerns internally were ignored. 

Krupski told the BBC “I don’t think the hardware is ready and the software is ready. It affects all of us because we are essentially experiments on public roads. So even if you don’t have a Tesla, your children still walk on the footpath.”

Tesla has not yet responded to Mr Krupski’s claims.

Apple is rolling out the first 17.3 beta to developers this week, with the inclusion of a ‘Stolen Device Protection’ feature.

In the unfortunate event that your iPhone and passcode are stolen, this new added layer of security aims to provide peace of mind by using biometric information, like a face scan or fingerprint, to access critical phone functions.

In this article, we’ll explore exactly what Stolen Device Protection is and how you can switch it on. However, if you’re unable to access the 17.3 beta just yet, don’t worry! The latest 17.2 update has also just landed.

Setting Up Stolen Device Protection

Stolen Device Protection activates when your phone is in an unknown location. As soon as this happens, it’ll kick in and share a notification immediately, as well as a second alert one hour later.

Those who have updated to the iOS 17.3 beta will have seen a prompt to test a preview of the feature following installation. However, Apple has said this prompt won’t pop up for those who install iOS 17.3’s public version.

The feature is opt-in and super simple to set up. Here’s how:

1. Head to the Settings

2. Scroll to Face ID & Passcode

3. Then Stolen Device Protection

4. Toggle it on

Warning: Downloading beta versions of iOS can be problematic, as new features aren’t fully tested, you also run the risk of losing data, and losing functionality. Apple itself recommends not downloading beta versions to ‘business critical’ devices. If you want to give the beta a spin, we’d strongly recommend downloading it to a secondary iPhone.

The Benefits of Stolen Protection

When the feature is switched on, particularly sensitive actions will need Face ID or Touch ID authentication in order to be used. They’ll also feature a one-hour security delay, in which a user must wait 60 minutes before authenticating again.

However, this one-hour delay won’t be necessary if your iPhone is being used in a familiar location, like your home or workplace.

Here are the actions it’ll help to protect:

• Changing your iPhone passcode
• Adding or removing Face ID or Touch ID
• Changing your Apple ID password
• Updating select Apple ID account security settings, including adding or removing a trusted device, trusted phone number, Recovery Key, or Recovery Contact
• Turning off Find My 
• Turning off Stolen Device Protection

Several actions will also require the same biometric authentication in order to be used, but won’t feature the one-hour delay. These include:

• Deleting all content and settings
• Viewing an Apple Card virtual card
• Applying for a new Apple Card
• Using payment methods that are saved in Safari
• Viewing or using passwords or passkeys saved in iCloud Keychain
• Turning off Lost Mode
• Making certain Apple Cash and Savings actions in Wallet
• Using your iPhone to set up a new device

New Feature Follows Theft Exposé

This new security feature comes following a report from The Wall Street Journal earlier this year, which exposed how the recovery key was being hijacked by criminals and resulted in victims permanently losing treasured family photos and other important digital items. 

Reporters Joanna Stern and Nicole Nguyen discussed instances of thieves spying on their victim’s passcode before stealing the device. They were then able to turn off Find My, reset the Apple ID password, and view passwords for banking and email accounts that were stored in iCloud Keychain.

Stolen Device Protection will be available on all iPhone models that are compatible with iOS 17. It’s likely the full public rollout will take place in January or February 2024.

Ernst & Young LLP has this week released its third annual EY Future Workplace Index, a national research survey that provides insight into the C-suite of US companies. 

From utilizing AI within the workplace to redefining Casual Fridays, the survey is chock-full of interesting data and analytics. However, a stand-out finding shows that there’s been a remarkable shift away from fully remote working in the last 12 months.

These statistics are at odds with the research that companies with remote working policies show higher revenue growth, or that 4-day workweeks and flexible working benefits both employers and employees.

A Shift From Working Fully Remote

It’s no secret that the world of corporate real estate is still finding its feet after three years of unsteady ground, with hybrid, remote, flexible, and four-day week work models all being tried and tested by companies.

However, the EY Future Workplace Index has shown that the wildly different stances seem to be settling down in favour of back-to-the-office mandates. 

EY US survey respondents reported that full-time remote working has plummeted from 34% in 2022 to just 1% in 2023. 

In the same vein, 99% of respondents indicated that they required employees to be in the office at least two days a week, with 32% of them requiring three days a week. Only 1% of employees are being asked to go into the office once a week or less. 

Is Corporate Real Estate Growth Inevitable?

While these findings may disappoint employees who have been able to better balance work and life priorities thanks to more autonomy and flexibility, they do provide a glimmer of home for those in corporate real estate.

The drop in fully remote work highlights a likely chance that office towers and parks will start coming back to life. This in itself will have a positive economic impact for not only the real estate agents, but the retail and hospitality businesses in surrounding office areas.

But that’s not all. According to Francisco J. Acoba, EY Americas Co-Lead for Corporate Real Estate Consulting and Technology, more people in the office may boost employee productivity too.

“Dark offices can create unintended doubts and uncertainty for both the workforce and the marketplace, even while all that cost is still being absorbed by corporate budgets. We’ve learned a lot in the past three years about productive ways of working, and it’s becoming clear that most workplace models benefit from keeping the office lights on.”

Around 80% of companies reported that their employees’ productivity had been “somewhat or much higher” over the past 24 months. In contrast, around 28% of leaders felt their employees were equally productive when they work from home.

More Survey Highlights

While the battle of home versus office working rages on, the Index results also provided interesting insights elsewhere.

The introduction of AI in the workplace is being embraced, albeit with caution. 44% of respondents said they are using AI to “collect data to maximize and optimize their office space”. This is largely to create an efficient balance of individual and collaborative areas that target specific user needs.

Similarly, 38% stated they want to “apply AI to track the office’s sustainability and energy efficiency”. Only 2% of respondents said they’re not using AI or are not sure.

Elsewhere, talk on backtracking on Monday-Thursday working weeks took centre stage, with the percentage of employers implementing a four-day workweek falling from 21% to 18% since last year. Not all hope is lost though, as 24% of respondents are still considering the four-day workweek.

However, if you’re still firmly on the side of flexible working, even after reading the report and emerging trends, our tips on how to ask to work remotely may be helpful.

Snapchat has this week released its latest AI-powered feature that allows users to create and send AI-generated images from a text prompt.

Available for Snapchat+ subscribers only, the feature follows in the AI footsteps of the company’s Dreams feature – a tool that lets users create AI images for their chat wallpapers and bitmoji backgrounds.

Not prepared to fall behind in the AI innovation game, Snapchat’s release comes just days after Meta AI announced its text-prompt image generator, Imagine. 

How Does the New Snapchat AI Feature Work?

In a blog post shared this week, Snapchat explained exactly how the new feature works.

To start creating, subscribers must tap the ‘AI’ button on their right-hand menu. The feature will open with prompt examples – such as ‘a raccoon driving a motorcycle’ – which users are free to play with, as well a space to type in their own creations.

The generator will get to work and then deliver the image. Users will be given the option to edit, download, or share with their friends. 

As well as the image generator, Snapchat has also released an AI-powered extension tool. For photos that are zoomed in too close, users can tap on the ‘crop’ button, then the ‘extend’ option and AI will get to work filling in the external area to provide a broader, fuller picture. 

There’s currently no word on which AI model is behind the new features, but Snapchat has made it clear that the experience is “privacy centric”.

Other Snapchat AI Features

This isn’t Snapchat’s first foray into AI, with its Dreams generative-AI selfie feature having been available since August. 

With it, the instant messaging app encourages users to ‘try on new personas powered by AI’. This means transforming normal selfies into fantastical creations set in various scenarios, such as a deep-sea mermaid or renaissance-era royal. 

Users can create up to eight AI-generated selfies, with more available following an in-app purchase.

Similarly, Snapchat announced an expansion of its My AI feature back in April. My AI is an in-app AI-powered chatbot for users to ask questions and seek responses with, much like ChatGPT or Bard.

My AI is currently being used by over 750 million Snapchatters a month, with over 10 billion messages having been sent with it.

Snapchat Going All In on AI

The launch of its AI-image generator, in addition to the existing AI features, highlights Snapchat’s intent on going full-throttle with its artificial intelligence capabilities. 

With the likes of Meta’s Imagine and Google’s Gemini already available and making waves, in-app AI tools are becoming less of a trend and more of a necessity in order to keep up. The only difference being that these two are free for users, and Snapchat’s currently remains behind the Snapchat+ subscription paywall.

Snapchat+ launched in the summer and currently costs $3.99 a month. Right now it has more than 7 million subscribers , with subscription revenue having already increased by double digits in the majority of countries it’s available in.

The new Snapchat+ AI features are currently rolling out, with regional availability varying.

What’s next for Google Gemini, now that the internet giant has launched its new AI model? According to the latest reports, an ultra-intelligent AI assistant being developed under the codename Project Ellmann could be on the horizon.

Project Ellmann doesn’t sound so much like a ChatGPT rival as it does a potential ChatGPT killer. Rather than just answering your questions and helping you get mundane stuff done, the secretive Google initiative is being designed to capture a “birds-eye view” of your life.

The main offshoot of the undertaking would be Ellmann Chat, which Google imagines as “ChatGPT…[if] it knows everything about your life.” It’s a fascinating if slightly startling prospect, given the incredible amount of data and personal information Google already knows about most internet users.

What is Google Project Ellmann?

News of Project Ellmann’s existence come to us via CNBC, which claims to have seen documents related to the initiative used to present at a internal company summit. The presentation was apparently delivered by a product manager for Google Photos alongside representatives from its Gemini teams.

The AI assistant will ingest your personal photos, documents you may have stored in a Google account, and other contextual information available on the web so that it can effectively tell “your life story.” This will then be fed into a Gemini-powered chatbot, codenamed Ellmann Chat, and provide responses that are personalized to a currently unprecedented extent.

Examples given include deducing that you like pasta from the presence of Italian food in many of your photos, which sounds a bit like Google Photos’ current Memories feature doing a fireside chat with Big Brother. The end result won’t be limited to just better restaurants recommendations, either, with Google’s presentation claiming that Ellmann would ultimately be able to know when you were born, who your parents were, and if you have any siblings – even if it didn’t directly ingest these facts.

Is Gemini x Project Ellmann a Potential ChatGPT Killer?

If these reports are accurate, the budding ChatGPT vs Gemini rivalry could be set for many more new and dramatic twists in 2024 and beyond.

Google launched its latest AI large language model (LLM) earlier in the month to much initial fanfare, followed by some controversary after it was discovered the Gemini hands-on video was staged.

The latest version of Gemini is currently being rolled out to Bard, Google’s current publicly available AI chatbot, as well as being available for customers to license and develop their own applications.

Project Ellmann and Your Privacy

Of course, it will take a lot more than a backroom presentation for ChatGPT vs Bard to be decided once and for all, but news of Project Ellmann has likely raise a few eyebrows at its Microsoft-backed rival, OpenAI (which owns ChatGPT).

It’s also likely to raise eyebrows amongst the general public. Google gaining new AI superpowers to learn everything there is about, well, everyone isn’t likely to be a purely charitable endeavor.

As with the current stockpile of data Google holds on its users, knowing more about you ultimately serves to better inform not only what Google tells you, but what it shows you in terms of advertising and ecommerce. Project Ellmann is sure to set a few tinfoil hats on fire, given the company’s far from spotless track record in matters of personal privacy and the Wild West of AI ethics at present.

Apple users could be looking at a (small) payday, if they fit the requirements of an ongoing lawsuit against the major tech company.

Users who purchasd Apple gift cards are being encouraged to come forward, after complaints levelled at Apple that it was negligent with regards to the value and security of Apple gift cards.

It may not be the biggest tech settlement we’ve seen this year (that remains Facebook’s eyewatering $775 million pay out), but if you are eligible, you could receive a check from Apple.

What is the Apple Lawsuit About?

Apple was taken to court after a class action lawsuit alleged that the company didn’t do enough to safeguard its Apple gift cards, leaving buyers susceptible to loss in cases where they purchased cards that had already been redeemed.

Apple has denied these allegations, but has agreed to settle out of court, by stumping up $1.8 million for those that have been affected by this issue with its gift cards.

Apple gift cards have long been desirable in the scamming community, due to their easy accessibility and lack of paper trail. It has been alleged in the past that scammers were able to easily note down Apple gift code numbers in-store, redeem then, and then replace the card back on the shelf, leaving the legitimate purchaser with noting more than a plastic card with zero value.

How to Cash In On Apple’s $1.8 Million Settlement

Interested in claiming your entitlement? Read on to learn more about how to make a claim.

Am I eligible for Apple’s settlement?

If you’re a US resident that purchases an Apple gift card between March 2018 and July 2020, and your gift card had already had a redemption attempt prior to your purchase, you are eligble.

It’s worth noting that if you’re a resident of California, the dates are slightly different, running from May 2017 to February 2018.

How do I make a claim?

To fill out your claim online, go to the Settlement Website and click on “Submit Claim”.

You’ll need to provide some information alongside the form, so make sure to have the following to hand:

What information do I need to provide?

As part of the claim process, you’ll need to provide the following information:

• Your full name
• Your address
• Your email address
• Your phone number
• Date the Apple gift card was purchased
• Proof of purchase for the Apple gift card
• The payment type you’d prefer (e.g. check by mail, direct deposit etc)

What’s the deadline for Apple’s settlement?

The deadline to file a claim is on January 8th 2024. Claims filed after this date will not be processed.

How much money could I receive?

Usually we don’t know how much to expect in a class action settlement like this one, but in this case, it’s straightforward – you’ll receive the value of the original Apple gift card you purchased.

There is one caveat – if the total number of claims exceeds the $1.8 million compensation pot, claimants will receive payments on a pro rata basis.

When could I receive compensation?

The final approving hearing is on January 17th, so barring any last minute issues, claimants can expect to receive their payments some time after this.

ChatGPT’s pricing plans start with a free plan. There’s also a ChatGPT “Plus” plan for users who want the premium experience, which retails at $20 per month and includes access to OpenAI’s most advanced large language model, GPT-4.

ChatGPT also has products and services geared towards businesses, but all of them are priced by the token, rather than per user, per month like most other software is.

So, how much does ChatGPT cost, what’s on offer, and how does it compare to the pricing plans offered by other popular chatbots, such as Anthropic’s Claude? Read on to find out more in this ChatGPT pricing guide.

• ChatGPT Pricing Explained
• The Free Version of ChatGPT
• ChatGPT Plus Pricing
• ChatGPT Enterprise Pricing
• DALL-E Pricing (Images)
• OpenAI Whisper Pricing (Audio)
• Does ChatGPT Offer Discounts?
• ChatGPT Pricing vs Competitors

ChatGPT Pricing and Terms Explained

Before we dive into precisely how much ChatGPT costs, we’ll quickly go through how ChatGPT prices its products and services. ChatGPT’s pricing structure changes depending on what you’re using the chatbot for and the specific services/models you’re using.

If you already know how ChatGPT’s pricing works and you’d prefer to get stuck into ChatGPT’s costs, simply skip this section and continue reading from our rundown of the free version.

ChatGPT’s consumer offerings

ChatGPT has a free and premium version geared towards consumer use. These versions of ChatGPT can be used to assist individuals at work with a wide range of tasks, but they can’t be fine-tuned in the same way ChatGPT’s business-focused offerings can, and you can’t edit the training data either. Overall, they’re less customizable.

ChatGPT Enterprise (API)

ChatGPT Enterprise customers don’t pay per account, or even per user, for access to the API. Instead, they pay for “tokens”. When Large Language Models (LLMs) process text, it’s broken down into small units, called “tokens” – which might be entire words, but are typically bits of words.

Enterprise users pay for both input tokens (the information they provide) and output tokens (the responses that ChatGPT generates).

If you’ve been exploring what ChatGPT can do for your business, you’ve probably heard or seen the phrase “context window” (e.g. “GPT 3.5 Turbo has a 16K context window”). The “16K” – or 16,000 – in that sentence refers to the number of tokens

A context window is the maximum amount of text a language model can consider at any one time – which includes the prompt and the response it gives. If you input a prompt that takes a chatbot over its context window limit, it’ll start forgetting the start of any conversation you’ve had with it.

Check out our guide to the best ChatGPT prompts for writing to learn more

Image generation

Along with ChatGPT, OpenAI also has an image generator called DALL-E. Currently, there’s DALL-E 2, DALL-E 3 standard, and DALL-E 3 HD. From October 19, 2023, DALL-E 3 has been available to all ChatGPT Plus and Enterprise customers.

When using DALL-E, rather than paying for tokens, you’ll instead pay per image generated. The price-per-image will differ depending on the version that you’re using, because the more recent versions are much better at producing accurate imagery of a higher quality.

ChatGPT Free Version

As you’re probably aware by now, ChatGPT has a free version which was launched on November 30, 2022 – and as the previous sentence suggests, you don’t have to pay anything for it. To use it, all you have to do is sign up for an OpenAI account with an email address and a phone number.

ChatGPT’s free version is powered by GPT-3.5 Turbo. This is the latest of the GPT-3.5 class of LLMs. While this is a very advanced LLM, it is not OpenAI’s most advanced effort. In a GPT-3.5 vs GPT-4 head-to-head, GPT-4 comes out on top every time (despite not being as speedy).

Along with access to the chatbot, OpenAI account holders also have access to a free version of DALL-E (DALLE-2) which can generate images, although once more, this isn’t the most advanced version of the software currently available.

ChatGPT Plus Pricing

ChatGPT Plus, which was introduced in February 2023, currently retails at $20 per month. Plus users can use GPT-4, which is more powerful but takes longer to answer, as well as GPT 3.5, which is quicker but not quite as capable.

ChatGPT Plus will let you chat with images and recordings, and you’ll be able to create images without leaving the interface (in the free version, you’ll have to switch to DALL-E 2). You can also use DALLE-3, the most advanced version of the image generator, separately if you like.

Another big difference between the free and paid versions is that you can build your own chatbot using ChatGPT Plus, called GPTs. This feature is not available in the free version of chatGPT. You can also use other peoples’ GPTs, although the GPT store, which was scheduled to open this month, has been postponed to 2024.

ChatGPT Enterprise (API) Pricing

ChatGPT’s Enterprise API models and services are priced per 1,000 tokens. According to OpenAI, 1,000 tokens are roughly equivalent to 750 words.

GPT-4 Turbo pricing

GPT-4 Turbo is ChatGPT’s most powerful LLM and has a 128K context window, which is among the largest in the industry. At the moment, the per-token price is cheaper than it is for GPT-4.

GPT-4 pricing

GPT-4 is like GPT-4 Turbo’s incredibly similar little brother – they’re both very powerful, but crucially, GPT-4’s training data cuts off in September 2021, like GPT 3.5.

GPT-3.5 Turbo pricing

“GPT-3.5 Turbo,” OpenAI says, “is the flagship model of this family, supports a 16K context window and is optimized for dialog.” Along with the Turbo version, there’s an instruct model which supports a smaller context window of 4,000 tokens. As you can see, for both inputs and outputs, it’s a little cheaper than GPT-4 Turbo and GPT-4:

GPT base model pricing

The GPT base models offered by OpenAI are for businesses on a budget – they’re a lot more basic than the likes of GPT-3.5 and GPT-4 and, as OpenAI admits, they’re just not as good at instruction following. However, they’re more than capable of performing a narrow range of tasks if fine-tuned in the right way.

Fine-tuning model pricing

Businesses can create custom versions of ChatGPT by inputting their own training data. ChatGPT will then learn things from this training data and be able to give much more useful responses than if it was just relying on the general dataset used to train all GPT LLMs.

You won’t be charged for inputting training data on these plans, just the input and output tokens used in requests. As you can see, you can fine-tune GPT-3.5 Turbo or, if you can’t afford it right now, a more basic language model.

Assistants API pricing

ChatGPT’s Assistants API was announced at ChatGPT’s first DevDay, which took place in November. It essentially lets developers build their own, customized chatbots for businesses, which can live inside existing applications. It’s currently in beta, and OpenAI says they’re currently working to make it more functional. If you’re using the beta, OpenAI will only charge you for inputs at the moment:

DALL-E Pricing (Images)

As we’ve covered, OpenAI will charge you per image to use DALL-E, and the prices vary depending on both the model you’re using and the resolution of the images you’re generating. Here’s the pricing table on OpenAI’s website:

OpenAI Whisper Pricing (Audio)

OpenAI now provides an audio model called “Whisper”, which can convert plain text into audio speech/audio files. Here’s the current pricing as listed on OpenAI’s website:

Does ChatGPT Offer Discounts?

The short answer is “no”. While some software programs like monday.com offer discounts for startups and non-profits, OpenAI is currently not offering any discounts for any businesses. So, if you want access to the Plus or Enterprise capabilities, you’ll have to pay full price.

In business terms, this is quite understandable. The demand for ChatGPT is so high that OpenAI isn’t exactly crying out for new sign-ups, and besides, a free version is already available. It costs around $700,000 a day just to run the chatbot – so Sam Altman and Co. need every cent they can get.

ChatGPT Pricing vs Competitors

When ChatGPT launched in November 2022, it was virtually one of a kind. Now, there are several competitors on the market, and a couple of them outperform ChatGPT when it comes to some tasks while also keeping pace in other areas.

ChatGPT vs Claude

Claude 2 is the most recent LLM powering ChatGPT competitor Claude, a chatbot created by Anthropic. Anthropic is a startup like OpenAI and is financially backed by Amazon and Google, the latter of which owns a 10% stake in the company.

Claude is a highly capable language model and powers Jasper – it also has a 128K context window, meaning it can handle incredibly large inputs.

Like ChatGPT, there’s a free version of Claude, as well as a $20 per month premium version called Claude Pro. Claude also provides per-token pricing for businesses that want to use Claude 2 or 2.1, as well as a lightweight version called Claude Instant. As you can see from the pricing information below, Claude’s prices are set per million tokens, rather than per thousand:

For more information on Anthropic’s pride and joy, check out our guide to Claude.

ChatGPT vs Bard

Google’s ChatGPT competitor Bard was rushed to launch in February 2023 – around the same time that OpenAI launched ChatGPT Plus. Then, it was powered by the LaMDa large language model, which a Google employee once said displayed sentience.

More recently, it’s been shifted to PaLM 2, a more powerful model, which Google says is faster and more efficient than LaMDA. Soon, it’ll be making the switch to Gemini, which creators DeepMind have said is more powerful and intelligent than GPT-4. For now, however, GPT-4 Turbo is widely considered to be the most powerful language model available.

Despite its reputation for the odd mistake, Bard’s big advantage is that it’s completely free. All you need is a Gmail account (personal or business), which doesn’t cost anything to create, and you’ll have access to its full functionality. To see how the two chatbots stack up against each other in a series of tests, check out our Bard vs ChatGPT comparison.

Apple has emphatically shut down Beeper Mini, an application that allowed users to send iMessages – which are sent using Wi-Fi rather than cellular networks – between Android and iOS-powered devices.

On Friday, the company announced on social media platform X that they were experiencing an outage. It later transpired that this was due to Apple’s efforts to block it, with the tech giant citing “safety concerns”.

In this short guide, we’ll explain why Apple wanted to shut Beeper Mini down so badly, whether Beeper is safe, and whether any Beeper alternatives are out there are worth considering. All in all, we cover:

• What Is Beeper?
• Why Did Apple Shut Down Beeper?
• Is Beeper Safe? Security Measure Explained
• Is Beeper Fixed Yet?
• Sunbird: A Beeper Alternative to Avoid

What Is Beeper?

Beeper Mini is a mobile app that launched in early December 2023. The app allows Android users to send iMessages, a form of correspondence typically exclusive to iPhones, iPads, and other devices running iOS.

“Beeper”, on the other hand, is an all-in-one unified messaging app for iOS devices and desktops that was launched back in 2021. After the launch of Beeper Mini, it has been renamed “Beeper Cloud”.

Although it’s been possible to send iMessages without an iPhone before the existence of Beeper, the company’s method is a newer, simpler, and more secure way of doing this.

Beeper doesn’t even require users to create or hand over their Apple IDs.

Beeper Mini was launched after a security researcher – a high school student at the time – submitted a Python script that managed to reverse engineer Apple’s iMessage protocol and provide the functionality to Android phones.

Why Did Apple Shut Down Beeper?

Towards the tail end of last week, Reddit began to be filled with reports of a Beeper outage as users’ message requests were timing out en masse.

The reports were later validated by Beeper co-founder Eric Migicovsky on X (formerly Twitter), who said that “all data indicates that” Apple was behind the outage.

Apple explains in a recently released statement that it “took steps to protect our users by blocking techniques that exploit fake credentials to gain access to iMessage.”

“These techniques posed significant risks to user security and privacy” the company’s solitary statement to the press on the issue continues, “including the potential for metadata exposure and enabling unwanted messages, spam, and phishing attacks.”

Apple’s decision to block Beeper Mini provoked the ire of US Senator Elizabeth Warren, a long-term advocate for more stringent anti-trust laws, who accused Apple of “squashing competitors”.

Is Beeper Safe?

In short, Beeper Cloud has some security limitations that make it less secure than using encrypted messaging apps. However, Beeper Mini is a lot more secure – the company’s openness about its security architecture and rigorous app testing is encouraging, and you don’t need to hand over any Apple account information to use it.

Importantly, Beeper Mini seems like it’s safer than other Android-to-iMessage options at the moment, as well as sending unencrypted SMS messages. Security is rarely a zero-sum equation – and currently, Android messages aren’t encrypted. What’s more, there seems to be no obvious security issue with the way the app has been configured, despite Apple’s protestations.

Beeper Cloud

There are limitations to Beeper Cloud from a security perspective, that are worth knowing about. The company states in its “Getting Started Guide“:

“Beeper is a universal chat app that supports connections to 15+ chat networks. To use Beeper, you must give the app permission to send and receive messages through other chat networks using your account credentials. By definition, this may be less secure than using other chat apps alone, especially encrypted chat apps like Signal.”

On top of this, a security report pointed out in August, if users wanted to send messages via Beeper, they’d have to hand over their Apple account information – which is a “huge risk”.

However, this is not the case with the recently launched Beeper Mini (more on this below) which Apple has blocked – you don’t have to log in with any credentials.

Beeper Mini

Beeper Mini has gone to great lengths to make it as secure as possible for users, which is encouraging – and it’s so confident in its work that it’s open-sourced its code.

As we’ve said previously, it’s certainly safer for Android users than sending standard text messages – and your Apple account ID is not needed to use the app.

Beeper Mini encryption and security process explained

At present, Android/”text” messages (i.e. “green bubbles”) are unencrypted. This means they can be read by anyone who wants to – including your phone carrier. There are very few protections in place to stop this kind of intrusive behavior from happening.

In contrast, when you send a message from an Android device using Beeper Mini, it’ll be end-to-end encrypted (E2EE) before it is sent, and thus sent securely. It does this by implementing Apple’s E2EE protocol natively in the Android app.

“We built Beeper Mini by analyzing the traffic sent between the native iMessage app and Apple’s servers and rebuilding our own app that sends the same requests and understands the same responses,” a company blog post explains.

Beeper says it cannot view the contents of any messages sent using its app, and the private encryption keys used to secure messages – as well as contacts – don’t leave users’ local devices. Only public keys are sent to Apple’s servers, and messages aren’t transferred as plain text.

A diagram showing how Beeper’s message routing system works. Image: Beeper

Unlike other apps that provide similar services, Beeper Mini connects directly to Apple servers rather than using a Mac server relay, something that competitors have struggled to do thus far. This is considered more secure than a middleman company hosting its own servers.

In this way, Beeper Mini effectively tricks Apple into thinking that Android messages sent via its services are iMessages, which means it can take advantage of Apple’s Gateway security system.

However, considering this is the case, questions remain about how Apple was able to parse these messages from standard iMessages and stop Beeper Mini in its tracks last week.

Beeper Mini diagnostics and reporting services

Importantly, Beeper Mini uses very few additional services and apps for diagnostics and analytics reporting and lists them on its security blog.

The company utilizes “a self-hosted installation of Rudderstack for analytics and diagnostic events” – this is used for app improvements but can be disabled by users in settings. OneSignal and RevenueCat are also used, the company says.

Beeper Mini security testing

Beeper says that it performed a red team analysis on Bleeper Mini. That means it got a team of security experts together to attempt to breach the app like a threat actor would, to spot its weaknesses and vulnerabilities, and then subsequently fix them.

Encouragingly, the company invites independent researchers to contact the company if they would like to perform similar analyses and provides an email address for them to do so.

Is Beeper Fixed Yet?

While Beeper Mini isn’t up and running again just yet, Engadget reported over the weekend that the company says it’s “very close” to a fix for the current disruption. 

However, now that it’s confirmed that Apple is behind the outage rather than a technical issue or cyber-attack, the app might be down for a while longer.

In the meantime, many people will look for a Beeper alternative to continue to send iMessages from their Android devices – but users should be very careful about what they decide to use. 

Sunbird Messaging: a Beeper Alternative to Avoid

According to Google Trends data, in the wake of Beeper’s outage, an application called “Sunbird Messaging” has seen a sharp uptick in searches.

Sunbird messaging markets itself as a “unified inbox” application that, similarly to Beeper Cloud, aggregates messages across various platforms, and claims to include secure Android-to-iMessage support.

The 9to5Google reports that the app was made available as a closed alpha to those who signed up for its waitlist during 2022, and claimed to use E2EE similarly to Beeper Mini. However, the app was put on pause in late November 2023 due to “security concerns”, with several sources claiming that the app wasn’t end-to-end encrypted in the way specified.

Sunbird had partnered with Nothing – a phone brand company owned by OnePlus co-founder Charles Pei – to launch Nothing Chats, which was also pulled from app stores around the same time as Sunbird was paused.

We’d recommend avoiding these apps, and also bear in mind that any apps claiming to be either of these two services is also not to be messed with. The great lengths Beeper has gone to in order to ensure their app is secure illustrates how hard it is to produce this kind of technology in a secure fashion.