Researchers have found that thousands of servers, running AI infrastructure, have been hacked in an active attack campaign. This attack targets a reported vulnerability in Ray, a computing framework used by the likes of OpenAI, Uber, and Amazon.

According to analysts at Oligo, a cybersecurity firm, the vulnerability allows attackers to take over the companies’ computing power and leak sensitive data. The blog post stated that this flaw has been under active exploitation for the last seven months, affecting sectors like education, cryptocurrency, biopharma and more.

I discovered an active attack campaign targeting a vulnerability in Ray. Thousands of companies running AI infrastructure are exposed to the attack through a critical vulnerability that is under dispute and thus has no patch.

It all began with a client that was using Ray. pic.twitter.com/BrVr2p8yOu

— Avi (@avi_lum) March 26, 2024

They claim that a trove of sensitive information from compromised servers has been leaked, targeting entities such as OpenAI, Hugging Face, Stripe, and Slack, as well as cloud environments like Amazon’s AWS and Microsoft Azure.

Oligo reports that it discovered hundreds of compromised clusters, with hackers allegedly installing cryptocurrency miners on compromised infrastructure. The researchers explain that attackers choose to compromise these machines because they can obtain valuable sensitive information, and GPUs are very expensive and difficult to obtain.

The company reports that GPU on-demand prices on AWS can reach an annual cost of $858,480 per machine, which means the total amount of machines and computing power that might have been compromised is estimated to be worth almost a billion dollars. Attackers have also installed reverse shells, which are text-based interfaces that allow for remote server control.

In a statement, it continued: “When attackers get their hands on a Ray production cluster, it is a jackpot. Valuable company data plus remote code execution makes it easy to monetize attacks—all while remaining in the shadows, totally undetected (and, with static security tools, undetectable).

What is the Ray AI framework?

Ray, an open-source unified compute framework, simplifies the scaling of AI and Python workloads, including everything from reinforcement learning and deep learning to tuning and model serving.

These applications generally operate on large clusters of servers. A central dashboard serves as an interface for displaying and managing active tasks and applications. Among the programming interfaces accessible via this dashboard is the Jobs API. It enables users to dispatch a list of commands to the cluster through a straightforward HTTP request that doesn’t require authentication.

In November 2023, analysts from the security outlet Bishop Fox spotted a similar vulnerability in Ray, tracked as CVE-2023-48022. Bishop Fox senior consultant Berenice Flores Garcia wrote in a blog post: “In the default configuration, Ray does not enforce authentication. As a result, attackers may freely submit jobs, delete existing jobs, retrieve sensitive information, and exploit the other vulnerabilities described in this advisory. “

Anyscale response

A spokesperson for Anyscale, the developer behind Ray, confirmed to ReadWrite that there were a number of issues it was fixing. In a statement, they said: “In light of reports of malicious activity, we have moved quickly to provide tooling to allow users to verify proper configuration of their clusters to avoid accidental exposure.”

They also said the firm was providing a client-side script and server-side code and that it had pre-configured the defaults of the client-side script to reach out to a server it has set up, simplifying the process of determining whether or not ports are unexpectedly open.

Last October, the company had initially denied the reports stating that four of the five reported bugs had already been fixed in November. They disputed the term “vulnerability,” referring to it as a bug instead.

It added: “We recognize that reasonable minds can differ on this issue, and consequently have decided that, while we still do not believe that an organization should rely on isolation controls within Ray like authentication, there can be value in certain contexts in furtherance of a defense-in-depth strategy, and so we will implement this as a new feature in a future release.”

Featured image: DALL-E

The post Massive hack hits AI servers, exploits Ray framework vulnerability appeared first on ReadWrite.

Google has announced several new updates to its Maps and Search functions, aimed at assisting users with their travel plans, thanks to AI.

Users in more than 40 cities throughout the U.S. and Canada will be able to access lists of suggested dining and visitation spots via Google Maps. These algorithm-generated suggestions comprise three main lists, which include a “Trending” list updated weekly featuring places currently experiencing a surge in popularity, a “Top” list of locations known for their enduring popularity, and a “Gems” list highlighting lesser-known but worthy destinations.

Are you ready to revolutionize your projects with immersive 3D experiences?

It's easier than ever to transform how users interact with geospatial data, thanks to our new solution guides.

Discover the possibilities → https://t.co/09x6vTsI38 pic.twitter.com/DF3larv3RL

— Google Maps Platform (@GMapsPlatform) March 27, 2024

In addition to this, Google Maps will feature curated recommendation lists from top sources including Lonely Planet, The New York Times, The Infatuation, and OpenTable. At the same time, Google Maps on Android and iOS is introducing enhanced customization features for creating personal lists.

For the first time, users can rearrange the order of places within these lists and link to relevant posts on their social media channels. These customizable, shareable lists are useful for trip planning, keeping track of favorite locations, and more.

In the upcoming months, they will begin showcasing some of the best lists from the Maps community. Therefore, if someone has created a list they believe others will appreciate, they can nominate it by sending it to listnominations@google.com.

Google will also be updating its design, giving Maps a new look, including a cleaner home screen with fewer tabs, and new pin colors that make it easier to find places on the map. The tech giant also states that it is using AI to generate intelligent summaries of locations in Google Maps, such as reviews and user-uploaded photos. It will also be employing AI to enhance photos of food with details, including the specific name of the dish.

Google’s push for more AI

In October 2023, the California-based company announced Immersive View for cycling, driving, and walking routes, offering detailed previews of journeys powered by Google’s AI. Users can view photorealistic 3D simulations of their whole route, helping them prepare for each turn and avoid issues like heavy traffic.

The feature also integrates AI-based analyses for weather and traffic forecasts. Immersive View, which initially launched in July 2023 for places, was extended to routes in cities globally. The expansion shows routes in 3D, as well as augmented reality search capabilities and information on electric vehicle charging.

Featured image: Canva

The post Google enhances Maps with new AI travel features appeared first on ReadWrite.

The British Broadcasting Corporation (BBC) has decided not to incorporate artificial intelligence (AI) into the advertising of the iconic show Doctor Who.

We reported earlier this month that David Housden, a senior executive with the BBC had initially said that there is “a rich variety of content in the Whoniverse collection on iPlayer to test and learn with, and Doctor Who thematically lends itself to AI, which is a bonus.”

Now the fans have spoken and the broadcaster has had to file an official statement saying “We have no plans to do this again to promote Doctor Who.”

BBC backpedals on AI incorporation

The Doctor Who Companion site was “disgusted” at the attempt to incorporate the generative AI into any aspect of the show.

The site would post that “shortly after the news was published, outrage from fans resulted in the (AI announcement) article being deleted — replaced by a webpage which states “Sorry, we couldn’t find that page.”

Housden had stated “We’re going to take it one step at a time, starting simple and learning as we go. We have chosen to start with Doctor Who, as it is a joint content priority for both BBC Public Service UK and BBC Studios marketing teams.”

This was a marquee moment for both the established pop culture character, who has fought storied battles against AI and robots in comics, television, and radio dramas, and the BBC.

The cherished Tardis-traveling Doctor and their fanbase won in the end. In response to the complaints levied at the BBC, the broadcaster released a statement acknowledging the backlash, writing:

“We received complaints about reports that the BBC is exploring generative AI in Doctor Who promotion.”

“As part of a small trial, marketing teams used generative AI technology to help draft some text for two promotional emails and mobile notifications to highlight Doctor Who programming available on the BBC.

We followed all BBC editorial compliance processes and the final text was verified and signed off by a member of the marketing team before it was sent.”

Doctor Who returns to BBC One and BBC iPlayer in May 2024 for the UK and Disney+ for the rest of the world.

As we reported earlier this year, Doctor Who will also be making a debut in the world of Fortnite to possibly coincide with this new series of science fiction adventures, according to well-established gaming leakers.

Image: BBC.

The post BBC in need of Sonic Screwdriver to fix the Doctor Who AI breakdown appeared first on ReadWrite.

Amazon is reportedly gearing up to invest nearly $150 billion over the next 15 years in data centers. This substantial financial commitment will equip the cloud-computing giant with the necessary resources to manage a projected rise in demand for AI applications and various digital services.

Bloomberg reports that this investment is a strategic display of dominance, to preserve Amazon’s leading position in the cloud services sector. Amazon currently holds about twice the market share of its closest competitor, Microsoft Corp.

An Amazon spokesperson confirmed to ReadWrite that the figures were based on its recent infrastructure announcements found on its website.

However, Amazon Web Services experienced its slowest sales growth on record last year, as corporate clients reduced expenses and postponed upgrades. Now, as spending begins to rebound, Amazon is eagerly securing land and energy for its energy-intensive operations.

In the last two years, Bloomberg said its calculations indicate that Amazon has pledged $148 billion towards the building and operation of data centers globally. The company aims to expand its server farm locations in northern Virginia and Oregon, and venture into new areas such as Mississippi, Saudi Arabia, and Malaysia.

Despite the expansion, AWS saw a 2% decrease in its data center investments in 2023, marking its first reduction, even as Microsoft ramped up its expenditures by over 50%, as reported by Dell’Oro Group. However, Amazon’s Chief Financial Officer announced last month that there would be an uptick in capital investments this year to fuel AWS’s expansion, encompassing projects related to artificial intelligence.

“As we look forward to 2024, we anticipate capex to increase year over year, primarily driven by increased infrastructure capex to support growth of our AWS business, including additional investments in generative AI and large language models,” said CFO Brian Olsavsky.

Amazon’s plans for AI expansion

While Amazon’s expansion of its data centers aims to cater to the growing need for corporate services, its focus on sophisticated, high-cost chips will provide the substantial computing power needed for the predicted increase in generative AI.

Reports suggest that Amazon is developing proprietary tools to compete with OpenAI’s ChatGPT, and has developed partnerships with various companies to enhance its AI services using its servers. As a result, Amazon expects to generate AI-related revenue amounting to tens of billions of dollars.

Featured image: Canva / Web Summit Rio

The post Amazon to ‘invest $150bn in data centers’ for AI growth appeared first on ReadWrite.

StealthMole, a rising dark web data intelligence startup, has announced that it secured $7 million in a Series A funding round.

We’re incredibly excited to announce that StealthMole has just closed a US$7 million Series A funding round!

A huge thank you to all our investors, partners, customers, and our incredible team for their unwavering support that has allowed us to reach this significant milestone.…

— Fusion Intelligence Center @ StealthMole (@stealthmole_int) March 28, 2024

The Singapore-based outlet said in a blog post that funds will be channeled towards StealthMole’s expansion into new markets, and supporting its efforts to apply its technology to more commercial uses.

The funding round was led by Korea Investment Partners (KIP), a multinational joint venture between RHL Ventures, Penjana Kapital, and KB Investment called Hibiscus Fund, as well as Smilegate Investment.

The company says it traces criminals using 255 billion analyzed data points from the dark web, deep web, and various hidden sources while leveraging advanced AI. Through AI and machine learning, it reportedly collects and connects data from hidden digital sources. This, in turn, aids governments and law enforcement in early risk mitigation and criminal tracking and supports businesses in cyber incident response and prevention.

StealthMole founder Louis Hur stated: “StealthMole came about from a critical market gap I encountered while working in cybersecurity and white-hat hacking: a severe lack of data points and information networks specifically within Asia.”

He added that data leaks, anonymized transactions, and cybercrimes were on the rise, demanding a better understanding of digital threats.

In response to the funding, managing director Kim Min-Q said: “As organizations across the world continue to digitize, cybercrimes will not only become more rampant but also more advanced. StealthMole has shown tremendous agility in this regard.”

Who are StealthMole?

Founded in 2022, the company is co-led by Simon Choi, an experienced threat investigator and former adjunct professor with extensive experience in working with South Korea’s national police and intelligence agencies, and Hur, a specialist in enterprise IT security who also founded a cybersecurity firm in South Korea.

Startups in the cybersecurity sector are facing a daunting market environment, contending with decreased valuations and increasing pressure to sell while competing for vital funding and collaborations.

Featured image: DALL-E

The post StealthMole secures $7m funding for its AI-powered dark web intelligence firm appeared first on ReadWrite.

An artificial intelligence (AI) model has been developed to predict early mortality, among other things, but as interest in the app has risen, the founders have begun issuing stern warnings.

The fortune-telling AI bot, named Life2vec, uses the sequences of life events to predict the future. This is helped by several years’ worth of registry datasets from Denmark, where the business originates from.

Deep learning models analyze information related to health, education, income, occupation, address, and working hours to predict everything from major life events to personality nuances.

While some may say it sounds like a bleak way to gain insight into your future, the creators describe the tool as being “fun and innovative” on their website.

With its growing popularity, the researchers have issued a stark warning to people who are curious to test their fate.

A message has been added to the website stating: “But first a warning: We are aware of life2vec social media accounts, and there is at least one fraudulent website. We are not affiliated with these or any other entities that claim to use our technology.”

Concerns have arisen due to fraudsters saying people can try the calculator for free, but the researchers say the software cannot yet be accessed via the Internet.

“We are working on ways to share the model with the wider research communities, but as LLMs are known to potentially leak data, we have to do further research before we can do this.

“We have not yet studied how our results generalize to other countries/contexts, but are actively investigating this topic.”

Predictions of time of death from AI bot

The predictions follow on from answers to general questions such as: ‘Death within four years?’

While the team says this isn’t the sole focus, they “chose the topic of predicting death, because it is a problem so many people have worked on. (For example, due to insurance companies, and so on).”

Featured image: Photo by Aron Visuals on Unsplash

The post Life2vec: Warning from creators of AI that predicts your death appeared first on ReadWrite.

Apple is set to reveal its expansion into the world of artificial intelligence (AI) at the annual Worldwide Developers Conference, hosted at Apple Park from June 10 through June 14.

The iPhone maker announced the details of the event through a press release on the Apple Newsroom, stating the event will bring “insight into new tools, frameworks, and features.”

While AI hasn’t explicitly been mentioned in the release, it’s expected this will be heavily discussed as iOS 18 is rumored to involve the use of artificial intelligence.

Chief Executive Officer Tim Cook promised that Apple will ‘break new ground’ in AI this year, as others in the industry have already begun exploring this new technology.

The SVP of marketing at Apple Greg Joswiak teased the event further on X describing what’s to come as being “absolutely incredible!”

Mark your calendars for #WWDC24, June 10-14. It’s going to be Absolutely Incredible! pic.twitter.com/YIln5972ZD

— Greg Joswiak (@gregjoz) March 26, 2024

This year, the conference will provide developers the opportunity to watch the keynote, meet with Apple team members, and take part ‘in special activities.’

The latest iOS, iPadOS, macOS, watchOS, tvOS, and visionOS advancements will be discussed too. Developers are said to see “unique access” this year, to help them elevate their apps and games.

Apple’s vice president of Worldwide Developer Relations Susan Prescott said: “We’re so excited to connect with developers from around the world for an extraordinary week of technology and community at WWDC24.

“WWDC is all about sharing new ideas and providing our amazing developers with innovative tools and resources to help them make something even more wonderful.”

Students to play a role in Worldwide Developers Conference

Alongside experienced developers, the next generation will be present as winners of the Swift Student Challenge will be eligible to apply for the in-person experience at Apple Park.

Thousands of people have been given the opportunity to showcase their creativity and coding capabilities through app playgrounds within the challenge which will recognize 50 distinguished winners.

Featured Image: Via Apple press release.

The post Apple set to unveil their AI master plan at Worldwide Developers Conference appeared first on ReadWrite.

Elon Musk’s AI chatbot, Grok, is set to be rolled out to all premium X users starting from the end of this week.

Musk himself took to the social media platform on Tuesday (Mar. 26) to share the news, with the AI tool only previously being accessible to Premium+ subscribers who pay $168 per year for the social network.

Later this week, Grok will be enabled for all premium subscribers (not just premium+) https://t.co/4u9lbLwe23

— Elon Musk (@elonmusk) March 26, 2024

Users can ask the chatbot any burning questions or queries and the answer can be directly posted onto X. An instructional video has been shared to give insight into how this will work and look alongside the Tesla-owner’s announcement post.

This news comes amidst an escalating rivalry with OpenAI and an ongoing lawsuit. Elon Musk is suing OpenAI and its CEO Sam Altman, as well as others, accusing them of violating contractual commitments established during the company’s inception in 2015.

The entrepreneur helped to co-found the ChatGPT parent company but left in 2018 citing a conflict of interest with the AI in his business.

Now, his attention has turned to his startup xAI which has since developed the Grok software.

What is Grok AI?

Grok AI was introduced by Elon Musk in November 2023 and it quickly became hailed as a substitute for ChatGPT – except it’s only available via the X platform.

Unlike other artificial intelligence technology, Grok has been programmed to exhibit a politically incorrect and witty personality, drawing inspiration from conservative commentators like Tucker Carlson and Joe Rogan.

It has real-time knowledge of current events too, including developments in the world of AI technology. It can craft stories, summarize website content, translate language, and research and answer questions.

Historically, it’s only been available to X subscribers on the most exclusive and expensive tier. This is described as a ‘special group’ on the app website.

It’s not yet known when Grok AI will be available to everyone, with the team focusing on development throughout.

The post Grok AI chatbot being rolled out for X premium users this week appeared first on ReadWrite.

The hotly anticipated AI assistant Rabbit R1 is set to begin shipping to US buyers this week.

According to Rabbit, the first batch of Rabbit R1 pre-orders will begin shipping on Sunday, March 31. With US deliveries taking a few weeks the company estimates that the initial R1 orders will be in customers’ hands ‘around April 24th’.

Debuting this year, the sherbet-toned orange Rabbit R1 has generated a lot of hype. Branded as an AI assistant, its arrival raises a lot of questions about the utility and impact of AI hardware in the age of smartphones. The Rabbit R1 is powered in large by San Francisco startup Perplexity’s LLM and the sleek and vibrant design by Teenage Engineering features a physical scroll wheel and a rotating camera, topped off with a cute animated bunny that serves as an avatar for the AI software.

The tech company started taking pre-orders of the Rabbit R1 for $199 in January – with the initial batches of 10,000 devices selling out rapidly. Batch one is set to release for shipping soon, however, buyers who purchased a later batch pre-order will have to wait a few months for theirs to arrive. Pre-orders for batch seven are currently live.

Rabbit is also hosting a pickup party in New York City at the end of April, open to all kinds of buyers and not just bulk orders. More information and details about the party and first deliveries will be released next week.

How does the Rabbit R1 work?

With such an appealing yet minimalist design, it raises the question of how this device works.

Well, the main way that users interact with the R1 is via pushing and holding the ‘Push to Talk’ button. This allows the Rabbit OS to start listening for input. An artfully designed rabbit head slowly rises and flows up and down as you speak your request. Booking a table for dinner? R1’s on it. Asking for a good Easter dinner recipe? No problem (Although no rabbit I hope). What track was sampled on Ice Cube’s hit ‘It Was A Good Day’? These are the types of questions that the Rabbit R1 can handle with ease, according to their controlled video demo.

Featured image: Rabbit: R1

The post Rabbit R1 starts shipping to the first batch of US buyers next week appeared first on ReadWrite.

Microsoft has announced the rollout of its Copilot artificial intelligence (AI) assistant across Windows 10 and 11 devices, but the new service does not appear to be operating with full efficiency at present.

An official update last week confirmed Microsoft Copilot was in operation, with the icon visible on the Windows taskbar but all isn’t quite as it seems.

The official blurb describes Copiot for Windows as “an AI-powered assistant that helps you get answers and inspirations from across the web, supports creativity and collaboration, and helps you focus on the task at hand.”

Although the AI assistant is embedded into Windows 11 version 23H2 devices by default, it is disabled on Windows 11 version 22H2 devices.

Further to that, Deskmodder users have uncovered additional access on other Windows 11 devices but crucially, the app is not powering the actual Copilot functionality and range of attributes as it stands.

How to check if Microsoft Copilot is available on my device?

The app may not appear at first glance but once Microsoft Edge is updated to the latest version, it will appear.

To check if the Microsoft Copilot tool is in use on your system, use the following path.

Go to Settings > Apps > Installed Apps.
Change the sort order to “Date installed” or type Copilot into the search field.
If you see Microsoft Copilot listed, it is installed.

You may also want to remove the app from your device with the uninstall option. Navigating to “advanced options” will reveal further details.

More information will be expected, but the findings from Deskmodder appear to indicate the Copilot app as a placeholder in its current form.

Users may also have a different experience, depending on their region, but the overall offering is unclear. Microsoft could later introduce a fully developed app to all devices with an update at some point, whilst there are also murmurs of work on Windows 12 which could be related.

Microsoft has been known to introduce new apps onto devices without a full launch or extensive information so we will have to watch this space.

Image credit: Microsoft

The post Microsoft Copilot arriving on Windows but still awaiting full functionality appeared first on ReadWrite.

The makers of ChatGPT, OpenAI, have announced their latest artificial intelligence engine, which is for creating videos from text prompts.

OpenAI already has Dall-E for generating static images, and now it has moved into videos to help complete its suite of AI tools for all creative purposes.

Sora AI is not yet readily available to use by the public. Still, it has been made available to a select few whose job it is to test the security and stability of AI products, also known as ‘red teamers,’ in February 2024.

But it is already causing a lot of talk on social media, with Sam Altman, the founder of OpenAI, demonstrating Sora’s abilities by replying to people’s prompt requests with the final product, generated by the video generator.

The early signs are that it is equally as impressive as ChatGPT and Dall-E, and it undoubtedly represents a new era of text to prompt filmmaking.

What is Sora AI?

Just as we’ve become so used to generative AI models for text and still images, such as ChatGPT, Dall-E and Google Gemini, Sora AI does the same but for videos.

The AI works in the same way as other generative AI models. It constantly learns from what it sees and consumes, and is trained to provide the most accurate and detailed response to whatever prompt it is responding to.

Sora AI is no different. If you give it a text prompt, for example, “blue boat sailing on the ocean with the sun beaming down”, you will get back a video of just that. You can be as specific or as vague as you like, but the more detail you give the AI model, the better your results will be.

Sam Altman’s Sora AI demonstrations on X in February will give you the best idea of how it works and just how accurate it is with enough detail in the text prompts.

For example, the video below is the result of the prompt, ‘A instructional cooking session for homemade gnocchi hosted by a grandmother social media influencer set in a rustic Tuscan country kitchen with cinematic lighting.’

https://t.co/rmk9zI0oqO pic.twitter.com/WanFKOzdIw

— Sam Altman (@sama) February 15, 2024

How does Sora AI work?

The technology behind Sora AI is the same as the one that lets you search for things on the internet. The more examples the AI sees, the better it will be able to spit the same thing in other images. Eventually, once an AI has seen enough of one thing, it will be able to generate its own version when prompted.

That is, of course, a very simplified way of explaining how generative AI works, but OpenAI has previously provided a more detailed explanation of how its AI model works. Sora AI is trained on publically available and licensed data to know what videos look like at a realistic level. It’s trained to know what it’s looking at and then uses that information to learn how to generate its own versions.

If you ask Sora AI to produce a video of a dog, it will generate results based on all the videos of dogs it has previously seen. It uses visual patches and building blocks to help understand what elements of the video should go where frame by frame. The more it sees and learns, the better and more accurate it becomes.

Sora’s technology is built on a diffusion model, where the AI begins with a messy response before working to improve its output through a series of feedback loops. It also uses transformer technology to use a number of data analysis methods to process vast levels of data, before learning to understand what important parts of the video to keep in, and the least important details to leave out.

What can Sora AI do?

To date, Sora AI is able to produce HD videos of up to a minute in length from text prompts. It is capable of generating ‘real world’, cartoon and CGI-style videos, but currently cannot include audio.

Sora AI can also generate videos from still images, fill in missing frames in existing videos and also stitch multiple videos together. It also has the capability of producing endless loops.

There also examples of it producing simulations of video games, such as Minecraft.

There are reportedly plans to add audio and editing tools to Sora AI, with the latter giving creators the ability to manually fix errors in the AI’s videos. Presumably, the AI will then learn from the manual fixes being made, but that is just a presumption and nothing more at this stage.

There are, however, some limitations, which can be expected. OpenAI has acknowledged these limitations, including people disappearing or transforming into other objects, as well as people moving in ways that just aren’t possible in the real world. OpenAI has confirmed they are already working on fixing these issues.

Once those issues are ironed out and Sora AI grows in sophistication and accuracy, it’s not entirely out of the question that we will be able to use it to create full-blown imaginary worlds, movies and even explore real places in the world without ever needing to visit them physically.

How can you access Sora AI?

Sora AI is not yet available to the public without an invite. Individual creators and testers are being asked to use the AI model and test it so OpenAI can act on feedback to ensure it’s ready for public release.

There is also the very important aspect of security and ethics. Generative AI models have been exploited by criminals and pranksters in the past – such as the sexually explicit deepfakes of Taylor Swift  – so OpenAI is putting a lot of effort into ensuring Sora AI can only be used for good and creative purposes.

That means you will not be able to use Sora AI to generate videos that show extreme violence, sexual content, hateful imagery or celebrity likeness. OpenAI will also include metadata in Sora AI to indicate its videos were generated by AI.

In an interview with The Wall Street Journal, OpenAI’s Chief Technology Officer, Mira Murati, suggested Sora AI will follow the same prompt policies as Dall-E, which means it will refuse to create videos of public figures.

It’s not yet known when exactly Sora AI will be made available to the public, but Murati did hint that it could be nearer the end of 2024. Meanwhile, there are some assumptions that when it does, it will be available as a web app and will eventually include additional features akin to ChatGPT, such as custom bots.

Featured Image: OpenAI // Sora AI

The post Sora AI: What is it? How to access video generator appeared first on ReadWrite.

The generative AI chatbot market is rapidly growing and while OpenAI’s ChatGPT might remain the most mainstream, there are many others on the market competing to be the very best for the general public, creatives businesses and anyone else looking to see how artificial intelligence can improve their day-to-day lives.

But which one is the best? ChatGPT may have been the first to go mainstream, but is it the market leader? Which companies have entered the generative AI chatbot space with a product worthy of taking on OpenAI’s offering?

Arguably the most popular on the market, other than ChatGPT, are Microsoft’s CoPilot, Claude by Anthropic and Gemini, which is owned by Google.

Here we look at all four of these popular generative AI chatbots and consider which one is the best for certain uses.

Key features of ChatGPT

At this point who hasn’t heard of ChatGPT? It was the first AI to go completely mainstream and show just how powerful AI can be to the wider public. It made such a splash, it reached one million active users within weeks of launching and now has over 180 million users worldwide and counting.

It’s creator, OpenAI, has worked tirelessly to keep it at the forefront of the market by launching new and improved features, including a Pro Version (GPT-4), web browsing capabilities and image generation, powered by Dall-E. There’s even the option to create your custom-made GPT-powered bot on any subject you want.

The free version, GPT-3.5, is only trained on human-created data up to January 2022, so it’s restrictive if you’re looking to use it for more up-to-date purposes involving real-time information. However, the Pro version, GPT-4, is available for $20 a month and is trained with data up to April 2023. Although that’s still relatively time-restrictive, it does also have access to the internet.

Is ChatGPT good?

Yes, at most taks, although it has had its controversies due to inaccuracies and misinformation, such as lawyers using it for case research and the chatbot fabricating historic cases. However, it remains a good first port of call for anyone just looking for an easy-to-use AI chatbot. It should be noted GPT-4 is significantly more effective than GPT-3.5, but the former is only available to paying users.

Key features of Microsoft CoPilot

CoPilot is Microsoft’s own generative AI chatbot, originating initially as a chat option on their search engine, Bing. It is now a stand-alone AI chatbot and is naturally built into all of Microsoft’s productivity and business tools, such as Windows and Microsoft 365.

Interestingly, Microsoft is a key investor in OpenAI’s ChatGPT, which was used to launch Bing Chat. GPT-4 continues to power CoPilot today and, like ChatGPT, also uses Dall-E to generate images.

That might sound like it’s no different to ChatGPT but Microsoft’s key USP with CoPilot is that it is ingested into all of the Microsoft tools and products billions of people use around the world every single day.

It behaves as an assistant to those who rely on the likes of Microsoft Excel, Microsoft Word and other 365 platforms to perform day-to-day tasks.

Is Microsfot CoPilot good?

The clue is in the name, but CoPilot is good for people who need help when using Microsoft’s extensive suite of tools, products, and software. It essentially behaves as an assistant, or co-pilot, inside these products.

From spreadsheets, text documents to computer code, CoPilot can help create it all with natural language prompts. Coders on the Microsoft-owned Github find it to be a very popular AI chatbot to use.

Key features of Google Gemini

Formerly called Bard, Gemini is owned by Google is another generative AI chatbot that is improving rapidly over time to rival GPT-4.

One major plus to Gemini is that it has no limit to the number of responses it can give you, unlike GPT-4 and CoPilot, which both have limits in this area.

That means you can essentially have long discussions with Google Gemini to find the information you require. On top of that, and rather unsurprisingly, Gemini bakes in a lot of the elements we’re all so used to from Google’s search engine. For example, if you ask it to help you plan a trip to a specific country, it will likely provide you with a map of that destination, using Google Maps, and may even dip into Google images to give you some kind of visual representation of the information it’s giving you.

Users can also add extensions, akin to Chrome extensions, for use in tools such as YouTube, Maps and Workspace.

Is Google Gemini good?

If you’re a big fan of Google products and apps, Gemini is likely the generative AI chatbot for you, but it’s also perfect if you’re looking for speedy interactions and unlimited prompts.

That’s because, while it isn’t faster than GPT-4, it has generally been found to be faster than CoPilot and GPT-3.5. But it’s not flawless and was recently caught up in controversy over the accuracy of its image generator amid claims it was ‘woke’.

Key features of Claude

The creators of Claude, Anthropic, is an AI company started by former OpenAI employees.

It’s something of an all-rounder, being a multi-modal chatbot with text, voice and document capabilities.

But the main praise it has had since its launch in early 2023 is the fluency of the conversations it can hold, its ability to understand the nuances in the ways humans communicate and its ability to refuse to generate harmful or unethical content, instead often suggesting alternative ways to accomplish what users are asking of it without breaking its own guidelines.

Claude recently launched Claude 3, which is a family of AI chatbots (Opus, Sonnet and Haiku) that offer varying levels of sophistication depending on what users require, and Anthropic claim its most powerful AI in the family, Opus, is almost 87% trained to undergraduate levels of knowledge and accuracy and 95% common knowledge and accuracy.

Is Claude good?

Claude’s extensive and powerful capabilities, such as being able to rapidly read, analyze and summarize uploaded files, make it a very useful generative AI chatbot for professionals.

It is also trained on real-time data, which undoubtedly speaks to Anthropic’s impressive claims of accuracy and levels of knowledge.

On Claude’s website, Anthropic claims it is a ‘next-generation AI assistant built for work and trained to be safe, accurate and secure.”

Featured Image: Ideogram

The post What is the best generative AI chatbot? ChatGPT, Copilot, Gemini and Claude compared appeared first on ReadWrite.

OpenAI’s revolutionary artificial intelligence chatbot, ChatGPT, has opened a whole market of similar generative AI models.

It has also led to some of the biggest tech companies in the world joining the party, including Google’s Gemini. However, one in particular is becoming the most competitive of them all, and that’s Claude 3.

Its creators, Google-backed Anthropic, recently launched Claude 3, a trio of language models catering to different user needs and preferences.

It’s similar to market leader ChatGPT, which offers GPT-3.5 as a free-to-access product, whereas GPT-4 requires a monthly subscription and for access to more features and a more sophisticated language model. We’re also expecting GPT-5 to launch soon.

The Claude 3 family is the same. Three language models, all with varying levels of detail and sophistication. One is free, the other two require a subscription.

What is Claude 3 and what can it do?

The generative AI Claude family includes three models – Haiku, Sonnet and Opus. According to Anthropic, the state-of-the-art models offer ‘increasingly powerful performance, allowing users to select the optimal balance of intelligence, speed and cost for their specific application.’

Today we're releasing Claude 3 Haiku, the fastest and most affordable model in its intelligence class.

Haiku is now available in the API and on https://t.co/uLbS2JNczH for Claude Pro subscribers. pic.twitter.com/uy25DFujdI

— Anthropic (@AnthropicAI) March 13, 2024

Opus is the most intelligent model and ‘outperforms’ its peers on most of the common evaluation benchmarks for AI systems. Anthropic’s website claims it has undergraduate-level expert knowledge and graduate-level expert reasoning, as well as being capable of performing basic mathematics. The AI company claims it ‘exhibits near-human levels of comprehension and fluency on complex tasks, leading the frontier of general intelligence.’

All models in the Claude family, though, display impressive abilities in analysis, forecasting, content creation, code generation and communicating in a variety of languages, including Spanish, Japanese and French.

Haiku is the fastest and most cost-effective model on the current market for its intelligence category, according to Anthropic. It is capable of reading and making sense of data and graph-dense research papers in under three seconds.

Meanwhile, Sonnet is ideal for tasks that require instant responses, such as knowledge retrieval or sales automation.

Anthropic says the Claude 3 family is capable of powering live customer chats, auto-completion and data extraction tasks, and can do all of the above in real-time. Meanwhile, all Claude 3 models are sophisticated enough to process a wide range of visual formats, including photos, graphs and diagrams.

How good is Claude 3?

Claude 3 uses a large set of complex, factual questions that are aimed at targeting known weaknesses in current models. The responses are categorized into correct answers, incorrect answers and admissions of uncertainty, where the model isn’t sure of the answer.

In Claude 3, Anthropic have been able to heavily improve the accuracy of their models compared to previous versions. It claims to have made a two-fold improvement in accuracy on challenging open-ended questions and reduced incorrect answers compared to Claude 2.1.

In a table on Anthropic’s website, each of the Claude 3 family is compared to GPT-4 AND GPT-3.5, and the results show far greater accuracy where Claude 3 is concerned.

Haiku, the least intelligent and costly member of the family, shows significantly higher accuracy levels over GPT-3.5, including in undergraduate-level knowledge, grade school math, coding, reasoning over text, knowledge Q&A and mixed evaluations.

The most intelligent and expensive of the trio, Opus, shows the same higher levels of accuracy over GPT-4, which is OpenAI’s equivalent. In fact, Opus is so accurate and sophisticated that it realized that it was being tested earlier this year, which was revealed by Anthropic prompt engineer, Alex Albert, on X.

How much does Claude 3 cost?

Claude on its own is free to use with usage limitations, while there’s a $20 monthly subscription for Claude Pro.

The Claude 3 API has three pricing levels that meet the levels of the three members of its family.

Haiku,  described as ‘Light & Fast’ by Anthropic, is priced at an Input of $0.25 per million tokens and an output of $1.25 per million tokens.

The ‘Hard-wording’ Sonnet comes in at $3 per million tokens for input and $15 per million tokens for output.

Finally, the ‘Powerful’ Opus is priced at $15 per million tokens for input and $75 per million tokens for output.

Is Claude 3 better than ChatGPT?

Anthropic make a range of claims that its Claude 3 family is far better than OpenAI’s GPT offering, and the accuracy results suggest they have a solid case.

Tech.co recently did a detailed test between the two to determine which one comes up trumps, asking them 13 different questions designed to test reasoning, natural language processing, ethics, creativity, brainstorming and spreadsheet formulas.
The results were quite clear-cut. The analysts said Claude 3 produced better responses in seven of the tests, while ChatGPT won three. There were four ties, where the responses were pretty much the same.

Claude 3 provided better, more detailed, and better sophistication in its responses to ethical reasoning, creating product descriptions,  some of its brainstorming, summarizing text, analyzing text, providing factual information and writing poems.

ChatGPT came out on top in creating spreadsheet formulas, composing an email and creative writing. It was deemed a tie in some of the brainstorming ideas questions, understanding natural language, personal advice and riddles and reasoning.

Given the level of accuracy, it’s easy to see why many would prefer Claude 3 over ChatGPT, but it all depends on what one’s needs are. For simple or basic prompts, it’s likely you won’t see a great deal of difference in responses between the two.

However, Claude 3 appears to be far more capable of accurately handling more complex and detailed tasks that would be more fitting in commercial or educational settings.

Featured Image: Anthropic

The post What Is Claude 3? Is it better than ChatGPT? appeared first on ReadWrite.

Besides all the benefits artificial intelligence offers the world – assisting with complicated tasks, streamlining work processes and helping with creative jobs – it also creates issues when people submit poorly written content using AI, and that’s where AI content detectors come in.

AI detectors aim to help the public differentiate between content created by humans and a computer.

As AI tools continue to get more advanced and more sophisticated, AI detectors need to do the same by ensuring they can accurately spot AI-generated content.

What is an AI content detector?

An AI content detector analyzes content to identify patterns in text, images or videos to decipher whether it is human or from generative AI. 

In the case of text, AI content detectors predominantly look for the repetition of certain words and phrases, a lack of depth, minimal creativity and inaccurate information.

For images and videos, they scan objects within the content to help determine their original source.

Similar to AI tools themselves, AI detectors are built on very powerful technological platforms and are designed to be just as sophisticated as the tools they’re analyzing.

The 5 best AI content detectors

There are loads of AI detectors on the market but not all are built the same and, more importantly, none are completely 100% accurate.

Furthermore, they all come with different perks, functionalities, pros, cons and price points. Some AI detectors are built for specific uses, others claim to do it all. But what are the best ones on the market right now?

We’ve listed our favorite five below…

Undetectable.ai

While some AI detectors seek to deliver an AI vs Human result when analyzing content, Undetectable.ai seeks to dig deeper into the details of the text it’s given and looks at syntax, style and structure.

It’s very easy to use. Simply paste the text you want the tool to analyze and click ‘Check for AI.’ It will then work its magic from there. It recognizes text from ChatGPT-3, GPT-4, Claude and Gemini, as well as other mainstream AI generators.

Undetectable.ai is careful not to guarantee any level of accuracy but it does claim to be the most advanced and accurate AI detection tool on the market, although third-party tests estimate around 85-95% detection rates. Moreover, it claims to be ideal for writers, bloggers and content creators.

How much does Undetectable.ai cost?

Undetectable.ai offers a freemium version that is limited to small-scale exploration. Personal plans with more flexible access to key features start at $5-a-month for 10,000 words, but with that you have to pay a year upfront.

For a monthly subscription, pricing starts at $14.99-a-month for 15,000 words.

Winston AI

Winston AI is aimed at businesses that want help detecting AI-generated content in marketing, advertising and other corporate information.

The cloud-based tool uses machine learning to help detect AI-generated content.

It is a user-friendly tool that comes with a range of key features, including file upload capabilities, OCR technology, printable PDF reports, plagiarism checks, project and document management and team management.

Users simply either paste their text or upload their file into Winston AI and the tool will then analyze the text before delivering a report to indicate whether it is AI-generated or created by a human. A neat feature is Winston color codes the full submitted text in the results into green, yellow and red depending how likely it is the content is AI-generated. Text highlighted in red means the application thinks there is a high chance it is AI-generated.

Winston AI claims to be up to 99.6% accurate, which is mightily impressive and precise. However, some third-party tests rate it closer to 85%.

How much does Winston AI cost?

Winston AI offers a variety of pricing plans that suit the needs of most users. They offer a free plan that comes with the most basic of features and an allowance of only 2,000 words, while the annual plan costs $12-per-month for 80,000 words. Monthly, you can expect to part ways with $18 for 200,000 words.

There is also an Elite plan for $49-a-month, which comes with 500,000 words of scanning allowance.

Copyleaks

Copyleaks is more sophisticated than other text-based AI detectors because it understands 30 different languages, so it’s the perfect tool for businesses with international interests. It also has the ability to analyze source code, which opens itself up to be incredibly useful for software developers.

The tool offers sentence-level analysis, which means the results you get from it will give you very precise details on which parts of your text were created by a human and which parts are AI-generated. On Copyleaks’ website, it claims users have “full transparency around the presence of AI-generated content even if the text has been interspersed with human-written content.”

Copyleaks claims to have 99.1% accuracy, but the very slight downside to using the tool is that it can be quite laggy. It does, though, detect the most popular AI models on the market, including ChatGPT and Gemini.

How much does Copyleaks cost?

Copyleaks doesn’t offer a free plan and its pricing tiers are based on credits, where 1 credit is worth 250 words. You can test it for free, though.

The cheapest monthly plan available is $10.99 for just 100 credits, but that only gets you access to Copyleaks’ AI Content Director but not their ‘award-winning’ plagiarism detector. To get access to the two, it’s $16.99 for 100 credits. Monthly plans max out at 10,000 credits, but the price skyrockets to over $900.

The cheapest annual plan for both the AI and plagiarism detectors is $13.99-a-month for 1,200 credits. For 120,000 credits, an annual plan costs just over $750-a-month.

ZeroGPT

If you’re looking for a very simple and easy-to-use AI detector, ZeroGPT offers just that. It supports various languages and has a very simple user interface that’s easy to follow and use.

Simply paste your text into ZeroGPT and it will analyze it for AI-generated content before giving you a percentage breakdown of how much of the text is written by a human and how much is AI or GPT generated.

It will highlight the areas of the text it believes are AI-generated in yellow, making it very easy for users to scan the text quickly.

ZeroGPT has API integration, which means the AI detector can be plugged into other apps or programs to detect AI-generated content in real-time, and also allows users to upload files, such as PDFs, to scan.

How much does ZeroGPT cost?

It’s free up to 15,000 characters per AI detection. If you need more than that, ZeroGPT’s Pro plan costs $7.99-a-month for 100,000 characters per AI detection and their Max plan gives you the same characters but more for the AI summarizer, AI paraphraser and AI grammar & spell check, which costs $18.99-a-month.

Crossplag

Crossplag was originally a popular plagiarism detection tool, and it still is, but it has also recently added its own AI content detector, which uses machine learning to identify whether text is AI-generated or written by a human.

Using patterns learned from human-written text, the tool is capable of recognizing when text deviates away from these patterns and will display a percentage of the chance of the text being AI-generated.

It is a remarkably quick and easy tool to use and is trained with over 1.5B parameters to give it the best possible chance of maximum accuracy.

How much does Crossplag cost?

Crossplag is free to use up to 10 credits, which is the equivalent to 1,000 words.. Beyond that, there is a Pay as you Go plan, which is 50 credits for €9.95, or there’s the ‘Bundle’ which is 1,000 credits for €149.95.

Anything beyond that and you need to contact Crossplag directly.

Featured Image: Generated by Ideogram

The post 5 best AI content detectors appeared first on ReadWrite.

Microsoft has unveiled a slew of AI-integrated products, including the new Surface Pro 10 and Surface Laptop 6, with its chatbot, Copilot, taking center stage.

The tech giant has been stepping up its game in the field of AI, with Copilot at the forefront of its innovations.

During Thursday’s livestream (March 21), Microsoft highlighted Copilot’s features, most notably its expanding capabilities in performing Windows system tasks for its users.

We’re excited to announce our first Microsoft @surface AI PCs built exclusively for business: Surface Pro 10 and Surface Laptop 6. Built with @intel® Core Ultra processors and @Windows 11—both bringing the Copilot experience to life: https://t.co/9GAYcuKjQw #Technology

— Microsoft 365 (@Microsoft365) March 21, 2024

Microsoft’s senior director for Windows Enterprise, Melissa Grant, said: “Windows 11 and Windows 365 promise a new era of AI productivity.

“Our goal is to empower employees to work more efficiently and creatively and make the most of their productive time.”

On Windows 11, users will be able to switch between “Work” and “Web” modes using the Copilot toggle. The Work button brings Copilot’s Microsoft 365 capabilities into the Windows interface.

Grant added that the aim of the new technology is to enable Windows to “deliver technology that puts you on a path to becoming an AI-centered organization.”

Microsoft is also investing in cloud PCs provided via Windows 365, using them as a platform for Copilot. The application used to access these cloud PCs is fittingly called the Windows app. Microsoft reports that since its preview release at Microsoft Ignite 2023 in November, the Windows App has accumulated over three million active usage hours.

The Washington-based corporation is optimistic that its Copilot AI companion will significantly save time and reduce stress in workplace environments.

Microsoft’s first AI PCs

The Surface Pro 10 is said to deliver the most advanced Copilot features in a portable design for mobile workers, whereas the Surface Laptop 6 combines the computational power typical of a desktop PC with the convenience of a lightweight laptop.

Users will get the chance to personalize the device for themselves, optimize power performance to extend battery life, and receive suggested performance enhancements and troubleshooting tips from Copilot.

The AI enhancements for Windows 365 will also help to increase device and organizational security through single sign-on, encryption of cloud PCs, and capabilities for zero-trust access.

The Pro 10 now uses a neural processing unit (NPU) that is the powerhouse behind all of the latest integrated Copilot AI features. This frees up capacity on the main processor to handle the tasks that matter.

It also features advanced inking technology, reportedly helping to organize notes into structured documents using Copilot. This comes with Copilot suggestions designed to refine workflow and enhance productivity.

The Pro 10 is said to offer the largest field-of-view camera on a tablet, providing 140 degrees of HD video angle along with AI-powered auto-framing.

Featured image: Microsoft / Canva

The post Microsoft focuses on Copilot as new AI Surface products revealed appeared first on ReadWrite.

Google DeepMind has officially launched a full AI system that focuses specifically on advising soccer coaches on corner kicks. TacticAI is described as being “an AI assistant for football tactics.”

Launched in collaboration with Liverpool FC, one of the best-performing teams in the UK, this is an artificial intelligence system that can provide experts with tactical insights through predictive and generative AI.

Announcing the news through a blog post on Google DeepMind yesterday (Mar. 19) the tool works by “using a geometric deep learning approach to help create more generalizable models.”

We're announcing TacticAI: an AI assistant capable of offering insights to football experts on corner kicks.

Developed with @LFC, it can help teams sample alternative player setups to evaluate possible outcomes, and achieves state-of-the-art results. … pic.twitter.com/8oVYtamGGe

— Google DeepMind (@GoogleDeepMind) March 19, 2024

This means teams worldwide could sample alternative player setups to evaluate possible outcomes to achieve the results they’re looking for.

To benchmark TacticAI, they have a dataset of 7176 corner kicks from the 2020 to 2021 Premier League seasons which are randomly shuffled and split into a training and a test set.

When in use, the tool is built to address three core questions. These include:

1. For a given corner kick tactical setup, what will happen? e.g., who is most likely to receive the ball, and will there be a shot attempt?

2. Once a setup has been played, can we understand what happened? e.g., have similar tactics worked well in the past?

3. How can we adjust the tactics to make a particular outcome happen? e.g., how should the defending players be repositioned to decrease the probability of shot attempts?

In a blind test with football experts, TacticAI’s suggestions were “indistinguishable” from real corners in previous games. When this group was asked to choose preferred tactical advice, they favored the AI’s ideas over the original 90% of the time.

Google and Liverpool FC’s multi-year partnership

Three years ago, Google announced a multi-year deal with the soccer club Liverpool F.C.

In their first joint paper from 2021, titled ‘Advancing sports analytics through AI research,’ soccer was then highlighted as being an interesting opportunity for AI. A year later, the Graph Imputer was created.

Graph Imputer showed how AI can be used with a prototype of a predictive system for downstream tasks in football analytics. The system was able to predict the movements of players off-camera when no tracking data was available.

The newly unveiled TacticAI is the third exploration into AI and sport by Google DeepMind and Liverpool FC.

Featured Image: Via Google DeepMind

The post Google’s DeepMind launches AI soccer coach appeared first on ReadWrite.

Artificial intelligence (AI) can help us do a lot nowadays, from brainstorming ideas or rewriting emails to creating whole artworks. One area that many people are finetuning more and more is AI character generation.

What is an AI character generator?

Just like any generative AI, an AI character generator uses prompts and machine learning to create a fresh image. Most apps use text-based prompts, taking information from the user and generating an image using that information.

Some generative AI tools will also use image-based prompts, where users upload several variations of what they want and the AI tool will combine them into a new image. This can be especially helpful to generate a stylized version of yourself, as you can add photos and ask for them to be remade in a pixel, graphic, or virtually any other style.

Best AI character generators

While all generative AI tools operate on the same principles, not all apps use the same AI algorithms. That means using different tools can result in different results, so it’s important to pick the right tool that fits your needs and preferences.

PhotoDirector

PhotoDirector’s main selling point is how easy it is to use. Available on both iOS and Android, it can turn photos of yourself into a wide range of diverse style, ranging from photorealistic to exaggerated graphic looks. You can also specify your gender expression to have results that match your identity.

Once an initial image is generated, it’s also easy to switch out details like hair color, style, and accessories to finetune the image as you please. 

FacePlay

FacePlay is a more artistic app, turning photos into artistic images from throughout time. From classical paintings to colourful anime or superhero portraits, any image can be transformed using the in-app tools.

The app also has a video function, which is more of a rarity in these sorts of apps. You can turn videos into cartoons, ready to share via messaging or on social media.

Dawn AI

While the first two options on this list work largely from image prompts, Dawn AI offers functionality for both text-to-image and image-to-image prompts. That means you don’t even need to have an original image to transform; you can simply write what you want to see and let the generative AI tool do the work.

There is also a range of AI styles to choose from, if you’re stuck for inspiration – as well as a library of prompts that’s expanding all the time. As well as being adept at creating characters, Dawn AI can also generate scenery, animals, and buildings.

Picsart

Standing out from the crowd of solo AI avatar tools, Picsart can generate fresh styles for couple or group pictures. While this applies to humans, there are plenty of examples of cartoon-style pet pictures with their owners, making it possibly the most adorable use of an AI tool yet.

With more than one million AI images generated every day, Picsart has one of the largest user bases on this list.

starryai

With starryai, the focus is on creativity and artistic expression, with the option to generate character avatars from images, text, and even emoji prompts. Unlimited prompts will cost you but you can try out a limited number of AI features and up to five characters on the free-to-use plan initially.

If you upgrade, more options to tweak AI images means that perfecting the avatar quickly becomes easier.

Fotor

Known for its photo-editing software, Fotor also has a generative AI app that’s ideal for character avatars. The text-to-image software has impressive accuracy, meaning you can rely on it to understand your initial prompt and learn fast when you correct it.

The built-in photo editing tools mean you can easily remove the background to create a quick PNG or edit other images to create collages, blended artworks, and more.

Artbreeder

Artbreeder prides itself on its collaborative tools. While it’s a web-based AI character generator rather than a mobile app, it has some of the best mixing and combining tools of any generative AI software. That means you can mix and match facial features, hairstyles, and clothing to create wholly new characters.

Character Generator

Character Generator does what it says on the tin – and it does it well. Developed by Adobe, this app’s USP is the ability to create 3D characters, ideal for use in games, movies, and other media. The emphasis is on photorealistic characters, complete with realistic facial expressions and clothing. There is also a wide variety of different body types, skin tones, and clothing styles, to finetune your avatars as closely as you can.

Lensa AI

Lensa AI is targeted towards influencers and promises to deliver a different result for everyone. It also offers guidance on the prompts you put in to help you get the results you’re looking for.

Once generated, you can touch up your images, add accessories or colors, remove objects, and make various other edits. It’s worth noting that the app does use your photos to train its AI, although it won’t save a record of them or share them with anyone else.

Featured image: Ideogram

The post 10 of the best AI character generator apps appeared first on ReadWrite.

The AI startup Inflection closed a $1.3 billion funding round in 2023, but just a year later its main investor Microsoft has taken on some of its key employees.

In a blog post published by Microsoft yesterday (Mar. 19) the CEO Satya Nadella shares a message to its employees:

“I want to share an exciting and important organizational update today. We are in Year 2 of the AI platform shift and must ensure we have the capability and capacity to boldly innovate…

“In that context, I’m very excited to announce that Mustafa Suleyman and Karén Simonyan are joining Microsoft to form a new organization called Microsoft AI, focused on advancing Copilot and our other consumer AI products and research.”

Mustafa Suleyman and Karén Simonyan were the co-founders of Inflection. They will now head up Microsoft’s newly formed AI division.

As for the rest of the company, “several members of the Inflection team have chosen to join Mustafa and Karén at Microsoft.”

California-based Inflection is centered around creating a conversational AI that you could talk to. Last May, it released Pi – a personal AI, designed to be empathetic, helpful, and safe.

The $1.3 billion funding round was led by Microsoft, Reid Hoffman, Bill Gates, Eric Schmidt and Nvidia.

Inflection publishes an update

Taking to their website, a blog post titled ‘The new Inflection: An important change to how we’ll work’ was published on the same day of the Microsoft release.

They announce the news of the co-founders leaving and say “We’re grateful for all their amazing work in getting Inflection to this stage, and wish them luck for this new chapter.”

A new CEO has been welcomed, called Sean White, and the third co-founder Reid Hoffman is described as continuing on the board and “excited to take these next steps in building personal intelligence for everyone.”

Looking forward, the AI team has said they plan to “lean into our AI studio business, where custom generative AI models are crafted, tested, and fine-tuned for commercial customers.”

Inflection 2.5 will now be hosted on Microsoft Azure.

Featured Image: Photo by Matthew Manuel on Unsplash

The post AI startup raises $1.3B, employees poached by its biggest investor, Microsoft appeared first on ReadWrite.

A new study that looks into the influence of artificial intelligence in the workplace has shown negative results, indicating a decline in quality of life while the use of AI rises.

The U.K.-based Institute for the Future of Work (IFOW) has looked into the link between worker wellbeing and technology exposure, with almost 5,000 people who are representative of the working population being analyzed.

Results have shown that “quality of life negatively correlated with frequency of interaction with newer workplace technologies such as wearables, robotics, AI and ML software.”

The team goes on to say in their published report that “this is consistent with research that connects such technologies to exacerbated feelings of disempowerment, increased sense of insecurity, task intensification and stress and loss of meaning, as well as anxiety and poorer overall health.”

Not all technologies have a negative impact though, with some actually helping with employee wellbeing.

“Results showed that digital information and communication technologies correlated with improved quality of life, whereas newer and more advanced technologies were correlated with reduced wellbeing.”

So computers and messaging tools have been proven within this research to bring more freedom and flexibility to workers, but the use of smart devices or AI isn’t as well received.

Use of AI in the workplace

One of the main focuses of businesses appears to be around the productivity of their employees, with the use of artificial intelligence being implemented to handle repetitive tasks or speed up processes.

Some are even opting to replace workers with AI completely, including the likes of tech giant IBM. Back in 2023, the CEO publicly announced plans to replace nearly 8,000 jobs with technology.

At the time, he noted that back-office functions, specifically in the human resources sector, would be the first to face these changes.

The post Workplace AI negatively impacts quality of life, study finds appeared first on ReadWrite.

It could soon be possible to speak without vocal cords, as engineers are focusing on a square-inch device that can help people regain their voice.

This could be extremely beneficial for people with pathological vocal cord conditions or those who are recovering from laryngeal cancer surgery as this tiny new patch is being developed by engineers at the University of California.

In a journal published in Nature Communications last week (Mar. 12), the researchers outlined their invention and the results of testing the device on eight adults.

They describe the wearable tool as being a “self-powered sensing-actuation system” based on soft magnetoelasticity to enable assisted speaking. It works by patients articulating sentences through muscle movements that are usually used in regular speech or lip-synching.

The sensing component of the device then recognizes the movements without the vibration of vocal folds. Electrical signals are then fed to a pre-trained machine-learning model that converts throat movement into voice signals.

Other solutions for helping people without vocal cords to speak are available but these largely include handheld electro-larynx devices or tracheoesophageal-puncture procedures. Both of these options could be considered to be inconvenient and uncomfortable.

How does it work?

The system was attached to the throat of the participants and data on laryngeal muscle movement was collected, with a machine-learning algorithm used to correlate the resulting signals to specific words.

The research team had the patients pronounce five sentences – both out loud and voicelessly – including ‘I don’t trust you’ and ‘I love you.’ Each person had to repeat the five sentences 100 times for data collection purposes.

The overall prediction accuracy of the model was 94.68%, working well across the different people involved.

Going forward, the research team plans to expand the vocabulary of the device through machine learning and test their findings in people with speech disorders. 

The post AI wearable device to help people speak without vocal cords appeared first on ReadWrite.