Inside the Google executive moves that led to its big AI reshuffle ↗
Google is tightening its grip on DeepMind as it works to put Gemini back out in front. Sergey Brin has reportedly been urging AI staff to go all-in on Gemini, including work around recursive self-improvement.
The reshuffle also gives chief AI architect Koray Kavukcuoglu more authority, while some teams are being folded more directly into Google. The slightly awkward part? Gemini's next flagship model has reportedly been delayed after trailing rivals in areas such as coding. (Reuters)
AI code review platform CodeRabbit valued at $1.5 billion in latest funding round ↗
CodeRabbit raised $143 million in a round co-led by Atomico and Smash Capital, lifting its valuation to $1.5 billion. The company says it's already carrying out more than 2 million code reviews each week for over 17,000 customers.
The timing makes sense: as AI writes more software, somebody - or perhaps another AI - needs to inspect the avalanche. CodeRabbit plans further expansion across Europe and Asia and says it will spend more than $10 million keeping its AI review tools free for open-source projects. (Reuters)
Natural-language coding startup Lovable raises $400 million at $13.3 billion valuation ↗
Lovable raised another $400 million, doubling its valuation to $13.3 billion. Menlo Ventures and EQT's Scaleup Europe Fund co-led the round, with Tencent and Balderton Capital among the other investors.
The natural-language app builder says annual recurring revenue has nearly tripled from $200 million and is tracking toward $600 million. Users have created more than 60 million projects - natural-language coding has gone from a niche experiment to a rather enormous business, or so it seems. (Reuters)
Cerebras shares plummet 16% after results fail to impress investors ↗
AI chipmaker Cerebras dropped 16% in extended trading after quarterly revenue missed Wall Street estimates, despite the company raising its full-year outlook. Sales reached $180.11 million, while analysts had expected $194.23 million.
Its cloud business roughly quadrupled to $126 million, though hardware sales declined. Cerebras is also scaling capacity for its huge $20 billion compute agreement with OpenAI - so the growth story remains very much alive; investors simply wanted it served hotter. (Reuters)
Cisco forecasts annual revenue above estimates on sustained AI spending ↗
Cisco forecast annual revenue above Wall Street expectations as demand for the networking gear behind AI clusters keeps climbing. The company recorded $9.3 billion of AI infrastructure orders during its latest fiscal year, including $4 billion in the final quarter alone.
Still, shares slipped after hours as investors picked over margins and already-lofty expectations. AI infrastructure is becoming an extraordinary money hose - but apparently even a money hose needs to spray faster these days. (Reuters)
Nebius powers past estimates as customers race to secure AI capacity ↗
AI cloud provider Nebius beat quarterly revenue expectations as companies rushed to lock down scarce compute capacity, sending its shares sharply higher. The value of its AI cloud deals nearly quadrupled, with four contracts averaging more than $1 billion each.
Nebius also raised its power-capacity target and said demand is strong enough that it could sell all of its planned capacity for the following year on current terms. Compute scarcity, somehow, continues to behave like beachfront property with GPUs attached. (Reuters)
American AI model makers smell an opportunity ↗
Cheap, capable Chinese open-weight models are forcing American AI companies to reconsider their strategies. Meta is returning to open models, while Nvidia is developing larger Nemotron systems aimed at competing with leading open-weight alternatives.
The economics are the kicker. Businesses increasingly don't need the absolute best frontier model for every task, making cheaper open models more attractive and putting extra pressure on expensive closed offerings from OpenAI and Anthropic. (Reuters)
FAQ
What does Google’s AI reshuffle mean for Gemini and DeepMind?
Google is bringing DeepMind more tightly into its broader AI strategy while giving chief AI architect Koray Kavukcuoglu greater authority. Sergey Brin has also reportedly encouraged teams to place a heavy focus on Gemini, including research into recursive self-improvement. The changes suggest Google wants closer, faster coordination as it works to strengthen Gemini against competing frontier models.
Why has Gemini’s next flagship AI model reportedly been delayed?
The next flagship Gemini model has reportedly faced delays after falling behind competitors in areas such as coding. Google’s executive reshuffle appears partly intended to improve coordination across AI research, product development, and leadership. A more integrated structure could help the company respond more quickly when model performance falls short of internal or competitive expectations.
Why are AI coding companies like CodeRabbit and Lovable attracting so much investment?
Investors are backing companies that serve different parts of the rapidly expanding AI software workflow. Lovable helps users build applications through natural-language instructions, while CodeRabbit focuses on reviewing both AI-generated and human-written code. As AI produces more software, tools for creating, checking, and maintaining that software can become increasingly valuable parts of the development process.
What is driving demand for AI infrastructure from Cisco and Nebius?
Companies are spending heavily on the networking, cloud capacity, and computing resources required to run increasingly large AI workloads. Cisco reported strong AI infrastructure orders, while Nebius said customers were racing to secure compute capacity through large cloud agreements. The figures suggest demand remains robust across both the physical networking layer and specialist AI cloud services.
Why did Cerebras shares fall despite strong demand for AI chips?
Cerebras shares dropped after quarterly revenue came in below Wall Street expectations, even as the company raised its full-year outlook. Its cloud business grew substantially, while hardware sales declined. The reaction shows that strong long-term demand for AI does not necessarily shield a company from short-term market disappointment when revenue or other closely watched results fall short of investor expectations.
Why are American AI companies becoming more interested in open-weight models?
Cheap, capable Chinese open-weight models are putting greater pressure on US AI companies to offer more competitive alternatives. Meta is returning to open models, while Nvidia is developing larger Nemotron systems. For many business tasks, users may prefer a lower-cost model that performs well enough rather than paying for the most powerful frontier model available.