GoKawiil
Tech News
Nvidia is defending its practice of financially backing AI labs and data-center builders, arguing that its core chip-sales business generates enough cash to sustain these commitments without strain. The company describes its revenue engine as a self-reinforcing 'flywheel' capable of funding both its own growth and the broader AI ecosystem it supports.
Risklytics, a member of YC's Summer 2026 batch, has launched as an insurance brokerage focused on companies deploying AI agents and robots in operational roles. The startup works with carriers willing to underwrite these emerging risks and manually reviews policy language before coverage is finalized, aiming to catch exclusions that standard commercial forms weren't written to address.
Companies building agentic AI systems have traditionally sent their internal data—tickets, customer records, contacts, and workflows—into third-party frontier models hosted elsewhere, relying on contracts for protection. A shift is emerging where successful enterprise AI deployments instead keep data within their own infrastructure and bring AI models to that data rather than exporting sensitive information externally.
Researchers from Google and the University of Illinois Urbana-Champaign tested large language models on variations of the classic Knights and Knaves puzzle, where truth-tellers and liars must be identified from their statements. The models frequently failed when puzzles were slightly altered from familiar training patterns, defaulting to memorized answers instead of reasoning through the new version. A similar pattern showed up on a benchmark called SimpleBench, where humans easily spot subtle twists that trip up even leading AI systems.
Thomson Reuters unveiled Thomson, its first proprietary large language model, which it built by fine-tuning an open-source base rather than training from scratch. The company says it spent roughly $40 million on the project, using its own legal, tax and news archives from Westlaw, Practical Law, Checkpoint and Reuters alongside hundreds of subject-matter experts to specialize the model for professional work.
Jeff Dean, who departed Google this month after 27 years including stints as head of Google AI and chief scientist, made his first public appearance since leaving at the 2026 Frontier & Pioneer Symposium. He advised young people entering tech to skim many research abstracts rather than deeply study a few, arguing this builds a broader mental map for connecting unrelated ideas. Dean, now CEO of startup DiscoveryLoop, also urged workers to pursue mid-length, roughly five-year problems rather than overly ambitious or overly simple ones.
A commentary argues that government contracts, grants and procurement rules for frontier technologies—space, AI and quantum computing—are quietly locking in dominant suppliers rather than fostering open competition. The authors point to the space sector, where repeated launch contracts to the same firms like SpaceX build data and performance advantages that make it harder for rivals to break in, and warn similar patterns are emerging in AI and quantum funding decisions.