It’s Friday
Here’s what’s worth your time today.
Meta has pushed back the release of its next foundational AI model, code-named Avocado, after internal tests showed it lagging behind top systems from rivals like OpenAI, Google, and Anthropic in areas such as reasoning, coding, and writing. The model had been expected to launch this month but is now delayed until at least May.
The setback comes despite Meta spending heavily to stay competitive in the AI race, including billions on hiring researchers and building data centers. The company reportedly discussed the possibility of temporarily licensing Google’s Gemini models to power some AI features while it continues improving its own system.
Supported by EnergyX
As AI infrastructure explodes, so does lithium demand. Google’s AI data centers currently use 100M+ lithium-ion cells for reliable, always-on power.
The problem? Big Tech’s race to build massive new data center capacity has Morgan Stanley forecasting an 80k-ton lithium shortage this year.
Few companies are better positioned to benefit than EnergyX. Their patented tech produces up to 3X more lithium than traditional methods at speeds 500X faster than evaporation ponds.
Beyond the tech, they have rights to ~150,000 acres of lithium-rich land across North and South America, including up to 9.8M tons of lithium in Chile alone.
Industry leaders like General Motors and POSCO have already invested. Join them as an early-stage EnergyX shareholder today.
Anthropic has introduced a new feature that lets its chatbot Claude generate interactive charts, diagrams, and visualizations directly inside conversations. The visuals appear inline with responses and can update or change as the discussion evolves, helping users explore concepts like data trends or scientific topics more interactively.
The feature is part of a broader push to make Claude’s responses more structured and useful, building on earlier updates that introduced formatted outputs for things like recipes and weather. Users can also connect Claude to tools such as Figma, Canva, and Slack to create and edit content directly within the chat.
Microsoft has filed a court brief supporting Anthropic in its legal challenge against the United States Department of Defense after the agency labeled the company a “supply chain risk,” a move that could block its AI from government systems. The dispute began after Anthropic refused to allow its AI assistant Claude to be used for mass surveillance or autonomous weapons. Tech companies including Google, Amazon, Apple and OpenAI have also backed Anthropic, arguing the designation could disrupt suppliers that rely on its technology.
A new report from ActivTrak suggests AI is adding pressure rather than reducing workloads. After analysing more than 10,000 workers before and after adopting AI tools, researchers found time spent on everyday tasks increased sharply, with email time doubling and messaging up 145%. Overall task time rose anywhere from 27% to 346%.
The shift is also cutting into deep-focus work. The average uninterrupted work session dropped by 9%, continuing a broader decline in concentrated work time. Researchers say AI is pushing employees to multitask more, filling their days with additional tasks rather than reducing them — a trend some experts warn could contribute to burnout and “AI brain fry” as workers juggle more decisions and tools.