The pulse, right now.
A continuously-updating read on what the AI industry is doing, shipping, buying, and saying.
appeal 9.0Meta Plans AI Compute Sales
Meta's reported plan to launch "Meta Compute" to sell AI compute and models, leveraging its $182.9 billion infrastructure investment, introduces a new hyperscale provider. This diversifies supply for procurement teams and platform engineers, potentially impacting the cost curve for AI development.
AI development is fragmenting, with specialised models and hardware emerging. Companies are monetising infrastructure, while governments increasingly scrutinise frontier AI releases. Concerns persist over AI's neutralit…
The stream
- RELEASEGemini Spark Google releases Gemini Spark for Mac7.0
- POLICYCloudflare Cloudflare blocks mixed AI crawlers by default7.0
- RESEARCHAnthropic, Sakana AI, Z.ai AI benchmarks show fragmented leadership8.0
- PRODUCTMeta Meta plans AI compute sales9.0
- INVESTMENTChipmakers Chipmakers add $2 trillion market cap9.0
- POLICYGodot Foundation Godot Foundation Forbids AI Code Contributions8.0
- FUNDINGEquiLibre Technologies EquiLibre raises $500M valuation in Series A9.0
- PRODUCTAnthropic Launches Claude Science Workbench for research9.0
- SECURITYAnthropic Claude Code Embeds hidden markers to track usage7.0
- PRODUCTOpenAI OpenAI unveils Codex hardware accessory8.0
- PRODUCTBase44 Base44 launches proprietary AI model Base18.0
- RESEARCHEuropean Space Agency AI accelerates astronomical discovery7.0
- + more stories in the last 24 hoursAll stories →
Subjects
AI development is fragmenting, with specialised models and hardware emerging.
A weekly synthesis of the most material AI-industry moves — what shipped, what was bought, what regulators said, and how the pattern is shifting.
Capital availability is no longer the primary constraint — return justification is.
Departments
All stories →
AI Benchmarks Show Fragmented Leadership
AI capabilities are fragmenting across specialised models and architectures, requiring architects and procurement teams to move beyond generalist benchmarks. Selecting the right model for specific tasks now dictates competitive advantage, shifting evaluation from raw scores to architectural fit.
Chipmakers Add $2 Trillion, AWS Expands FDE
Investors are broadening AI infrastructure bets, adding $2 trillion to Intel, Micron, and AMD's market cap. AWS's expanded forward-deployed engineering unit intensifies competition for tailored AI solutions, offering procurement teams more hardware options and challenging founders to assess bespoke integration services.
Godot Forbids AI Code Contributions
The Godot Foundation will prohibit AI-authored code and AI-submitted pull requests, citing maintainer burden and lack of AI accountability. This limits practical AI code generation use in open-source contributions, reinforcing human oversight for code quality and increasing review costs for AI-generated code.
EquiLibre Secures $500M Valuation for Trading AI
Consistent profitability from AI-driven trading models shifts expectations for automated financial strategies. EquiLibre's reinforcement learning agents, proven in poker, began trading on crypto markets in 2025 and now trade billions daily on S&P 500 and Nasdaq, claiming zero negative months since inception.
Anthropic Launches Claude Science Workbench
Anthropic's new Claude Science workbench centralises scientific research workflows, integrating databases and tools for enhanced reproducibility and reduced context switching. This approach, available in beta for Pro, Max, Team, and Enterprise subscribers, contrasts with competitors' gated, specialised models, shifting how research teams manage data and tools.
OpenAI Unveils Codex Hardware
OpenAI will launch a physical accessory for its Codex AI coding tool on July 15, developed with Work Louder. This macro pad offers dedicated hardware controls for developers, streamlining AI-assisted coding tasks and signalling OpenAI's focus on professional workflow integration.
Every Tuesday. 5-minute read. 34,828 unique readers.
A tight brief for CTOs, CFOs, product leads, and founders. No hype. No hedging. No filler.