Scientific discovery is driven by scientists generating novel hypotheses for complex problems that undergo rigorous experimental validation. To augment this process, we introduce Co-Scientist, a multi-agent AI system built on Gemini for structured scientific thinking and hypothesis generation. Co-Scientist aims to help scientists discover new original knowledge. Conditioned on their research objectives and prior scientific evidence, it formulates demonstrably novel research hypotheses for experimental verification. The system’s design involves agents continuously generating, critiquing and refining hypotheses accelerated by scaling test-time compute. Key contributions include: (1) a multi-agent architecture with an asynchronous task execution framework for flexible compute scaling; (2) a tournament evolution process for self-improving hypotheses generation. Automated evaluations show continued benefits of test-time compute scaling, improving hypothesis quality over time. While general purpose, we focus the validation in three biomedical applications: drug repurposing, novel target discovery ¹, and explaining mechanisms of anti-microbial resistance ². Specifically, Co-Scientist helped identify new drug repurposing candidates and synergistic combination therapies for acute myeloid leukemia, which were validated through in vitro experiments. These real-world validations demonstrate the potential of Co-Scientist to accelerate scientific discovery and usher in an era of AI empowered scientists. — Read More

#big7

Jassy was once Jeff Bezos’ deputy and the head of Amazon’s cloud computing arm. Five years into his tenure as CEO, he’s killing projects, cutting staff, pleasing Wall Street and steering the everything store through its greatest challenge yet.

… This July will mark five years since Andy Jassy took over the chief executive officer role from Amazon’s founder. At the corporate offices in Seattle, the workforce has grown accustomed to his brand of rigorous oversight and ongoing exhortations to act as if they were at Jeff Bezos’ startup, not a $2.9 trillion behemoth. He recently placed a series of staggeringly expensive bets on artificial intelligence, audacious even by the standards of Silicon Valley’s ongoing trillion-dollar AI bacchanalia. In February he agreed to invest as much as $50 billion in OpenAI in a deal that commits the rising startup to relying in part on Amazon’s data centers and custom-designed microchips. Then in April he expanded a similar partnership with its archrival, Anthropic—a $13 billion investment, with an option for an additional $20 billion. To Jassy’s critics, that spending was the price of Amazon’s late jump into the current AI wave. He wasn’t bluffing, though: Jassy spooked investors by vowing to spend $200 billion this year on big-ticket items including warehouse robots, a far-out effort to launch satellites into space, and in particular more AI data centers, AI chips and networking equipment. “I don’t think the world has ever seen a technology get this much adoption and grow this quickly, at least in my lifetime,” Jassy tells Bloomberg Businessweek. — Read More

Knowledge workers at Meta routinely contend with workflow fragmentation, where critical information — including meeting notes, tasks, key decisions, and code context — is siloed across disparate platforms. Each new AI conversation starts cold: the same explanations, the same links, the same ten minutes of context-setting before any real work begins.

So we tested a simple hypothesis: what if an AI agent had persistent, structured access to everything a person is working on, and carried that context across every interaction? Not a chatbot that answers questions, but a working partner that tracks projects, reads meeting notes, surfaces connections, and builds on prior conversations.
<brthat ai="" second="" brain="" experiment,="" born="" in="" the="" analytics="" org,="" has="" since="" been="" adopted="" by="" over="" 60,000="" people="" across="" meta:="" engineers,="" pms,="" designers,="" legal,="" finance,="" communications,="" and="" sales.="" this="" post="" covers="" how="" it="" was="" built,="" grew,="" what="" we="" learned.="" –="" Read More

#big7

Google is bringing conversational AI search to YouTube, marking the company’s latest push to infuse its products with AI-powered discovery tools. The feature, dubbed “Ask YouTube,” started rolling out to YouTube Premium subscribers in the US today as an experimental test. It transforms the platform’s search bar into a chatbot-style interface that pulls results from longform videos, Shorts, and text summaries – essentially giving YouTube its own version of Google’s AI Mode for search. — Read More

#big7

The race to secure electricity for AI models has reached new heights: Meta has signed an agreement with the startup Overview Energy that could see a thousand satellites beam infrared light to solar farms that power data centers at night.

In 2024, Meta’s data centers used more than 18,000 gigawatt-hours of electricity — roughly enough to power more than 1.7 million American homes for a year — and its need for compute power is only increasing. The company has committed to building 30 gigawatts of renewable power sources, with a focus on industrial-scale solar power plants.

Typically, data centers turning to solar power must either invest in battery storage or rely on other generation sources to operate at night.

Overview Energy, a four-year-old, Ashburn, Virginia, outfit that emerged from stealth in December, has a different solution: The company is developing spacecraft that collect plentiful solar power in space. It then plans to convert that energy to near-infrared light and beam it at sufficiently large solar farms — on the order of hundreds of megawatts — which can convert that light to electricity. — Read More

Meta released an AI model on Wednesday called Muse Spark, which marks its “first step” toward an “overhaul of [its] AI efforts.”

Muse Spark is the inaugural model to come out of Meta Superintelligence Labs, which was created last year because CEO Mark Zuckerberg was reportedly unhappy with the progress of Meta and its Llama models and how they lagged behind OpenAI’s ChatGPT and Anthropic’s Claude. Meta recruited former Scale AI co-founder and CEO Alexandr Wang to lead Meta Superintelligence Labs and invested $14.3 billion in the data labeling company for a 49% stake.

Now, it’s time for Zuckerberg to see if his reconfigured AI team can woo users. — Read More

[Google is] introducing a framework to measure progress toward AGI, and launching a Kaggle hackathon to build the relevant evaluations.

Artificial General Intelligence (AGI) has the potential to accelerate scientific discovery and help solve some of humanity’s most pressing problems. But it can be difficult to know how close we are to this key milestone, because there’s a lack of empirical tools for evaluating systems’ general intelligence. Tracking progress toward AGI will require a wide range of methods and approaches, and we believe cognitive science provides one important piece of the puzzle.

That’s why today, we’re releasing a new paper, “Measuring Progress Toward AGI: A Cognitive Taxonomy,” that presents a scientific foundation for understanding the cognitive capabilities of AI systems.

Alongside the paper, we are partnering with Kaggle to launch a hackathon, inviting the research community to help build the evaluations needed to put this framework into practice. — Read More

For years deep learning has followed one central idea. If we want smarter models, we stack more layers, run larger training, and scale everything upward. This simple formula has given us large language models that reason well and generate high-quality text. Yet they still share one huge weakness. They cannot learn on the fly. They cannot update themselves during use.

Any change needs heavy retraining, and this often destroys old knowledge.

Google Research recently published a paper called Nested Learning. It offers a very different way of thinking about how learning should work inside neural networks. The researchers claim that a model is not just a big stack of layers. It is a hierarchy of learners that operate at different timescales. If this view is correct, it could reshape how we build AI systems in the coming years. — Read More

Apple’s AI strategy has become a Rorschach test for the technology industry. Critics see a company falling dangerously behind. Needham analyst Laura Martin claims it is one to two years behind its competitors. But almost all of this commentary, whether bullish or bearish, focuses on the wrong question.

The standard narrative compares Apple’s AI capex to Microsoft’s, Apple’s Siri to Google’s Gemini, Apple’s foundation models to OpenAI’s GPT-4. By these metrics, Apple looks behind. But these comparisons assume Apple is trying to win the same race. The evidence suggests it isn’t. — Read More