Meta finalises Scale AI deal, valuing startup at $29 billion; recruits its CEO for 'superintelligence' team
Meta Platforms
has finalised an investment in Scale AI that values the startup at over $29 billion, Scale AI said on Thursday.
Two sources familiar with the matter said that Meta's investment in Scale AI amounts to $14.3 billion.
The sources said that Scale CEO and cofounder
Alexandr Wang
will join a new "superintelligence" unit inside Meta to achieve
artificial general intelligence
(AGI), a term that refers to machines that can match or surpass human capabilities.
by Taboola
by Taboola
Sponsored Links
Sponsored Links
Promoted Links
Promoted Links
You May Like
บ้านขนาดเล็กพร้อมห้องน้ำ ที่น่าสนใจในปี 2025
Visionary Echo
อ่านเพิ่มเติม
Undo
Scale's chief strategy officer, Jason Droege, will serve as its interim CEO, Scale AI added.
By poaching Wang, who does not have formal experience in frontier AI research, Meta CEO Mark Zuckerberg is betting that Meta's AI efforts can be turned around by an adept business leader more in the mold of OpenAI's Sam Altman than the research scientists at the helm of most competing labs.
Live Events
The deal ranks as Meta's second-largest ever after its $19 billion buyout of WhatsApp.
Discover the stories of your interest
Blockchain
5 Stories
Cyber-safety
7 Stories
Fintech
9 Stories
E-comm
9 Stories
ML
8 Stories
Edtech
6 Stories
Since many of those labs have contracted Scale for data services, Zuckerberg could also obtain an inside track into his rivals' priorities around data, one of the key ingredients for developing today's AI models.
Scale was valued at nearly $14 billion in a May 2024 funding round where the company raised $1 billion from backers including tech giants Nvidia, Amazon and Meta.
Founded in 2016, Scale AI provides vast amounts of accurately labelled data, which is pivotal for training sophisticated tools like OpenAI's ChatGPT.
To do so, Scale set up subsidiary platforms such as Remotasks and Outlier to recruit and manage gig-workers who manually label the data.
Wang, the 28-year-old co-founder, has steered the company to provide data labelling services across the buzzy tech sectors of the moment, from autonomous vehicles several years ago to generative AI today.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
4 hours ago
- Time of India
Meta AI may have a ‘personal problem' and a very serious one, and it's a warning for users
Representative image The Meta AI app may have a "personal chat problem" that has the potential to escalate into a major privacy issue. Users of the AI assistant developed by Facebook's parent company have complained that its "Discover" feed is reportedly displaying user prompts publicly without them being aware of the same. This feature was introduced with the transition from the Meta View app to the Meta AI app in April. It allows others to see the types of prompts people are submitting to Meta's AI chatbot. However, a concerned user named Justine Moore took to the social media platform X (earlier Twitter) to note observing prompts in the public feed that suggest users may not know that their queries are being openly displayed. This raises significant privacy implications for users interacting with the Meta AI service. How Meta AI is leaking personal chats of users In the X post, Moore shared screenshots of personal chats that Meta AI is showing other users and wrote: 'Wild things are happening on Meta's AI app. The feed is almost entirely boomers who seem to have no idea their conversations with the chatbot are posted publicly. They get pretty personal (see second pic, which I anonymized).' by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Free Solitaire. No Ads Ever. Play Solitaire Download Undo Later on, in the same thread shared transcribes of some chats, Moore further wrote: 'To clarify - conversations aren't public by default. You have to hit a 'share' button, but many users seem to think it's posting to a private journal. Resulting in things like this…a man trying to write a romantic poem for his gf. You can hear a 6 min audio clip, here's a transcribed excerpt:' 'Obsessed with this man who tries to use the app to find a woman with a 'big b**ty and nice r**k.' When it won't post on his behalf in local FB groups, he asks the bot to 'delete my number.' Instead, he (accidentally?) shares it publicly - I redacted,' Moore further added. As Moore suggests, users have been unintentionally exposing sensitive information on Meta AI. Users are advised to refrain from sharing prompts containing private medical and tax details, addresses, and intimate confessions (ranging from relationship doubts to personal dating inquiries) with the app as they seem to be appearing publicly. India's New AC Rule: Cooling Between 20°C–28°C Explained
New Indian Express
4 hours ago
- New Indian Express
Meta Unveils V-JEPA 2: A smarter AI that understands the physical world
Meta has launched V-JEPA 2, an advanced AI model trained on video, designed to help robots and AI systems better understand and predict how the physical world works. This model represents a big step toward Meta's long-term goal of developing Advanced Machine Intelligence (AMI)—AI that can think before it acts. Humans naturally understand how the world responds to actions—for example, knowing a ball will fall after being tossed, or avoiding people while walking through a crowd. This is because we build mental models of how things move and interact. V-JEPA 2 aims to give AI agents a similar understanding of their surroundings, helping them to observe, reason, and make better decisions before taking action. These types of AI systems are built using world models, which allow machines to understand, predict, and plan—just like humans do when navigating everyday situations. Smarter robots through video training V-JEPA 2 builds on Meta's earlier version of the model released last year, called V-JEPA. This new version offers major improvements in helping machines predict how objects move, how they interact with other objects, and how people handle them in the real world.
Economic Times
4 hours ago
- Economic Times
With Google's latest move, internet's becoming less ad-driven search, more user-paid GenAI
wwwhats-up@ As you browse the web, you must have encountered the 404 error, signifying that a web page can't be found. What you haven't seen is a similar code: 402. When Tim Berners- Lee was creating the World Wide Web, 402 was the code for 'payment required'. The original intention was that every visitor had to pay something to view a web page. But the schema was never built. So, there is no standardised encoded way to send or receive money online. In a 2019 podcast, Netscape builder Marc Andreessen explained: 'One would think the most obvious thing to do would be building in the browser the ability to actually spend money, right? You'll notice that didn't happen... we couldn't actually build economics into the core of the internet and so therefore advertising became the primary business model... We tried very hard to build payments into the browser. It was not possible... We made a huge mistake.' In 2014, internet expert Ethan Zuckerman wrote that 'advertising is the original sin of the web. The fallen state of our internet is a direct, if unintentional, consequence of choosing advertising as the default model to support online content and services.' Arguably, it's this 'original sin' that has made the largest destinations of the web the morass that it has become now. Thus, social networking has morphed into social media, as TikTok and Instagram pivoted to attention-seeking content that is optimised for views and clicks. This is also why Google's famous '10 blue links' are dominated by sponsored and advertiser preferences, rather than the succinct and accurate answer that a user wants. This drove even Berners-Lee to often regret his 'invention' and made him and others dream of a new kind of web - the concept of Web3, conceived to be 'owned' by users rather than advertisers, decentralised rather than concentrated with a handful of powerful tech companies, and, most importantly, have a different business model: that of micropayments. Thus, the user pays for the content she wants to consume, rather than advertiser paying for it, and the payment is also shared with the original the last 20 years, the company around which the existing web has revolved has been Google. While the 'advertiser pays' model was first brought in by companies like Yahoo and MSN, it was Google that perfected it as it made 'search' the way to organise the web. On the internet, if it's free, you're the product. Thus, Google and others sharply targeted and sold your attention, intentions and interest to advertisers who paid handsomely for it. How handsomely? Search contributed $54 bn of Google's $96 bn revenue in the last quarter, and reportedly most of its profitability. But it's also the model that held Google back and made it tiptoe tentatively into the world of GenAI. The search-through-chat business model of AI, which Perplexity or ChatGPT uses, does not lend itself as effectively to the advertising business model. Google experimented with 'AI Overviews' since I/O 2024. This model of AI-generated answers to common topics became popular with users, but reduced click rates on ads. This is the classical 'innovators' dilemma' that Harvard professor Clayton Christensen proposed in 1997, where he posited that great companies lose out to startup innovators, because they do not want to cannibalise their existing profitable business month, Google finally bit the bullet at I/O 2025, announcing that all its US users will be able to activate 'AI mode' in Google search and Chrome browser that will provide a conversational, question-and-answer experience akin to OpenAI's ChatGPT, rather than a traditional list of what happens to the super-profitable search business model? The company is tight-lipped about it, and is reportedly working to insert ads into AI answers. This is not as easy, however, since GenAI is a probabilistic tech, unlike the deterministic nature of traditional search. Thus, predicting what search result will come and matching advertiser needs to them becomes much more Google has hinted at a change in revenue model, with a $20 subscription for the service, and a $250 subscription for the super-premium version. So, with this monumental change, is the World Wide Web shifting on its axis? And does it take us one step closer to the dreams of a new web - user-owned, more democratic, and where you are finally not the product?Time will tell. But what is clear is that I/O 2025 did not signal just the re-emergence of Google in AI leaders, but also the emergence of a new kind of internet. So far, it was search around which the web was organised and the advertiser-driven model that paid for it. Maybe, we are now inching towards a web organised around GenAI, with subscribers paying for it - and content creators getting ably rewarded too. (Disclaimer: The opinions expressed in this column are that of the writer. The facts and opinions expressed here do not reflect the views of Elevate your knowledge and leadership skills at a cost cheaper than your daily tea. Warren Buffett-fan Pabrai is betting big on Edelweiss' Rashesh Shah. Will it pay off? Coal on one hand and green on the other; this company balances both Yet another battle over neem; this time it's a startup vs. Procter & Gamble Move over tariffs, China wields rare earths in an economic war of a different kind Is Zomato under siege? Quick commerce may be the next telecom 9 stocks from different segments of financial services sector with an upside potential of up to 37% Stock picks of the week: 5 stocks with consistent score improvement and return potential of more than 32% in 1 year Is an oil shock on its way? 14 stocks to watch carefully if the Iran-Israel conflict leads to a sustained rise in crude oil prices
