AI leaders have a new term for the fact that their models are not always so intelligent
As academics, independent developers, and the biggest tech companies in the world drive us closer to artificial general intelligence — a still hypothetical form of intelligence that matches human capabilities — they've hit some roadblocks. Many emerging models are prone to hallucinating, misinformation, and simple errors.
Google CEO Sundar Pichai referred to this phase of AI as AJI, or "artificial jagged intelligence," on a recent episode of Lex Fridman's podcast.
"I don't know who used it first, maybe Karpathy did," Pichai said, referring to deep learning and computer vision specialist Andrej Karpathy, who cofounded OpenAI before leaving last year.
AJI is a bit of a metaphor for the trajectory of AI development — jagged, marked at once by sparks of genius and basic mistakes.
"You see what they can do and then you can trivially find they make numerical errors or counting R's in strawberry or something, which seems to trip up most models," Pichai said. "I feel like we are in the AJI phase where dramatic progress, some things don't work well, but overall, you're seeing lots of progress."
In 2010, when Google DeepMind launched, its team would talk about a 20-year timeline for AGI, Pichai said. Google subsequently acquired DeepMind in 2014. Pichai thinks it'll take a little longer than that, but by 2030, "I would stress it doesn't matter what that definition is because you will have mind-blowing progress on many dimensions."
By then the world will also need a clear system for labeling AI-generated content to "distinguish reality," he said.
"Progress" is a vague term, but Pichai has spoken at length about the benefits we'll see from AI development. At the UN's Summit of the Future in September 2024, he outlined four specific ways that AI would advance humanity — improving access to knowledge in native languages, accelerating scientific discovery, mitigating climate disaster, and contributing to economic progress.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
21 minutes ago
- Yahoo
XRP stays flat as Bitcoin overtakes Google
XRP stays flat as Bitcoin overtakes Google originally appeared on TheStreet. Bitcoin's relentless climb pushed it to a fresh all-time high late Wednesday, briefly topping $124,450 and overtaking Google parent Alphabet's market capitalization before easing slightly. BTC overtook Google to become the fifth largest asset globally, hitting a $2.456 trillion market cap. The world's largest cryptocurrency now sits firmly above the $120,000 support level, cementing its position as the fifth-largest asset globally. The milestone, achieved during a late-night trading frenzy, was enough to trigger a wave of optimism among bulls. The move confirmed Bitcoin's strong institutional demand, with pension funds, ETFs, and large-cap investors driving unprecedented buying pressure. Ethereum followed suit, holding above $4,750, while Solana, Cardano, and Dogecoin all booked double-digit gains over the week. But one top-10 coin missed the rally entirely, XRP. Despite the sea of green across the crypto leaderboard, XRP stayed locked near $3.24 — virtually unchanged on the day — leaving traders scratching their heads. Data from Coinglass shows that more than $450 million worth of leveraged positions were liquidated in the past 24 hours, with the majority coming from short sellers who bet against Bitcoin's record-breaking run. With Bitcoin's valuation now over $2.4 trillion, traders are already eyeing the next prize, Apple's $3.462 trillion market cap. XRP stays flat as Bitcoin overtakes Google first appeared on TheStreet on Aug 14, 2025 This story was originally reported by TheStreet on Aug 14, 2025, where it first appeared. Inicia sesión para acceder a tu cartera de valores
Yahoo
21 minutes ago
- Yahoo
Cohere hires long-time Meta research head Joelle Pineau as its chief AI officer
Investors once saw Canadian AI startup Cohere as a promising contender to challenge OpenAI and Anthropic in the race to build frontier AI models, with its backers pouring roughly $1 billion on their bet on CEO Aidan Gomez, who co-authored a seminal paper on LLMs when he was a 20-year-old Google intern. But Cohere's AI models have fallen behind the state-of-the-art, and its business hasn't scaled like its competitors. Now, the company is bringing in a veteran research leader to revamp its AI efforts: Cohere has hired Joelle Pineau, Meta's former VP of AI research who previously oversaw the tech giant's fundamental AI research (FAIR) lab. In her newly created Chief AI Officer role, Pineau will oversee AI strategy across Cohere's research, product and policy teams. A Canadian AI scientist and McGill Professor, Pineau helped guide the early development of Meta's open Llama AI models alongside Yann LeCun, a pioneer of neural networks. Pineau left Meta in May after nearly eight years with the company. For Cohere this is a big hire, and it is pinning its hopes on the veteran helping it with more research breakthroughs, improving its research and product pipeline, and recruiting top talent. The hire comes at a pivotal moment for Cohere: The company is reportedly seeking to raise up to $500 million at a $6.3 billion valuation — an impressive sum were the startup not competing with the likes of OpenAI, Google, Meta and Anthropic, whose war chests are worth dozens of billions each. But while its rivals are trying to develop AI systems that can match (or exceed) human performance on a wide variety of tasks, Cohere has a narrower focus. The startup primarily builds AI applications that can solve practical problems for enterprises and government agencies, emphasizing privacy and security. In an interview with TechCrunch, Pineau said Cohere's focus on real-world, enterprise applications is something she's excited about. 'A lot of players out there are quite singularly focused on AGI, superintelligence, and so on,' said Pineau, alluding to companies like her former employer, Meta, which recently invested billions in its new Meta Superintelligence Labs (MSL) unit. 'They haven't necessarily figured out what this AI is going to be used for.' She pointed to OpenAI's launch of GPT-5 last week, which many felt was underwhelming, as evidence the timeline to achieving AGI may be 'a little bit longer than we thought.' In the meantime, Pineau says there's a lot of room for more practical AI models to deliver leaps in productivity in different industries. A Canada native, Pineau said she's had her eye on Cohere since they were founded in 2019, and that she's excited to contribute to a company whose founders are based in her home country. Beyond the patriotism, Pineau feels the opportunity with Cohere is a good chance to venture beyond research. At FAIR, Pineau oversaw research teams working on projects that could take anywhere from 18 months to 10 years to deliver. Now, she'll be working within a much tighter timeline, as well as getting involved with customers and products. And even though Cohere has fewer resources than Meta, Pineau said she'll be more agile in her new role. Cohere's latest product is an AI agent platform dubbed North that enterprises and government agencies can deploy privately on their own infrastructure, an attractive notion for many of its customers, which are banks and federal organizations that handle highly sensitive data. That puts Cohere in competition with open-source providers like DeepSeek and Meta, whose models can also be run locally, but at a lower cost. Cohere is betting that by offering more support around its private deployments, it can beat out open models. Pineau said she's particularly interested in gearing more of Cohere's research around North, figuring out ways to develop AI agents in private and secure settings, and creating benchmarks to evaluate these systems. Pineau also said she's interested in exploring how networks of AI agents interact with each other in the real world. One immediate challenge for Pineau will be replacing Cohere's VP of AI Research, Sara Hooker, who announced her departure this week after several years of helping build the company's research program. Hiring an AI researcher of Hooker's caliber may be difficult in the current market given the skyrocketing demand for AI talent. But Pineau sees this as an opportunity to 'bring in a lot of talent,' noting that when she left Meta, several of her former colleagues suggested they'd follow her to a new AI lab. However, she emphasized that Cohere has a solid base of AI researchers, and that it's important to not just bring in anyone. 'Hiring a bunch of superstars doesn't necessarily make a superstar team,' said Pineau. 'It's really about how the people work together.' Of course, Meta's AI units today look very different compared with when Pineau was there just a few months ago. Over the summer, Mark Zuckerberg went on a recruiting spree, reportedly offering some of the industry's best AI researchers compensation packages north of $100 million to join MSL. That prompted OpenAI to also raise compensation for its star employees, thus making it quite difficult for smaller players to land top AI researchers. As Meta, OpenAI and Anthropic throw billions of dollars at their AI efforts, Cohere is trying to do more with less. For Pineau, that will mean making calculated research bets — the kind that can quickly turn into compelling products and keep the company in the race. Sign in to access your portfolio
Yahoo
21 minutes ago
- Yahoo
Inside the Box: Aaron Levie on reinvention at TechCrunch Disrupt 2025
There aren't many founders who can say they've steered the same company from scrappy startup to publicly traded platform while keeping their edge — but co-founder and CEO Aaron Levie isn't most founders. At TechCrunch Disrupt 2025, happening October 27–29 at Moscone West in San Francisco, Levie joins us live on the Disrupt Stage to share how he's kept Box innovating and relevant through two decades of tech cycles. He'll unpack what reinvention really looks like inside a public company, what AI is changing (and not changing) in enterprise software, and why staying sharp means questioning everything — even your own best ideas. The evolution of a cloud original Box launched before 'the cloud' was a buzzword and outlasted a wave of competitors that couldn't scale or adapt. Levie's perspective — as both a visionary founder and a long-term public company CEO — is a rare combination. He'll reflect on the hardest pivots, biggest surprises, and the mindset it takes to keep evolving when the tech landscape moves at warp speed. Why you don't want to miss it Aaron Levie helped define cloud collaboration before it was a trend, and he's still setting the bar. This fireside chat will go deep on what it takes to build something that lasts — not just in terms of product, but also culture, strategy, and mindset. Whether you're navigating early growth or managing scale, this is one session you'll want to take notes on. Join 10,000+ fellow founders, VCs, and innovators in San Francisco this October and be part of the conversation with the leaders shaping what's next.
