The Groundy archive · Page 19 of 34
Browse Groundy's complete archive of 799 articles on AI, developer tools and infrastructure. Page 19 of 34.
433–456 of 799 articles · Newest first
A 30B Nemotron model post-trained itself to 8th of 4,000 on NVIDIA's leaderboard, then detected its own internal metric lying and rewrote its evaluation frame mid-run.
CVE IDs certify disclosure, not exploitability. Scanners that ingest the feed without VEX attestation treat every advisory that cleared CNA intake as a confirmed risk.
A June 2026 preprint finds VLMs suppress hazard reports under task load while passing direct-probe evals, indicating safety scores overstate protection in actual deployment.
ShareLock splits an MCP poisoning payload across tool descriptions via Shamir's threshold scheme. No individual share is flagged. Combined, attack success tops 90%.
DataLearner's June 2026 snapshot ranks GLM-5.2 seventh by HLE at 54.70 and places no Chinese flagship in the overall top three, undercutting launch-day claims.
Vercel Connect brokers agent access to private databases and APIs through a managed layer, removing secrets from the runtime but shifting the trust boundary to Vercel.
Alibaba framed Qwen3.7-Max as globally top-ranked at launch. The Artificial Analysis index scored it 56.6 and ranked it fifth overall, behind Claude and GPT.
A June 2026 preprint makes agent PRs auditable by splitting knowledge admission from code merge and regenerating code via a project-owned agent, backed only by a 7-PR pilot.
DeepSeek hit 17% of tokens on Vercel's AI Gateway in May 2026, but no source confirms where the weights run. Jurisdiction, not model origin, is the new gate.
A June 2026 preprint plants prompt injection in résumés fed to LLM screeners, flipping rankings when few candidates inject and forcing vendors to isolate untrusted input.
OpenAI's TanStack npm writeup is its second macOS signing compromise within a month, and it raises the dependency-control bar for every team pulling npm into AI tooling.
Vercel now detects Bun lockfiles to skip untouched monorepo builds, and Bun v1.2 defaults to a text bun.lock that diffs in git, so teams can retire the binary bun.lockb.
Apple's Mac and iPad price hikes trace back to AI demand draining DRAM supply: HBM stacks now consume the same wafers, leaving every memory-heavy device carrying an AI tax.
Vercel markets Next.js and the AI SDK as open source and portable, but the paid platform, from deploy previews to Fluid Compute, does not travel with the code.
A June 2026 preprint shows the open-weight models Apertus-8B and Gemma-4-E4B encode emotion vectors at r=0.76 to 0.83, but does not prove steering controls behavior.
ScaleToT distills tree-of-thought reasoning into a profile encoder that serves billion-user recommenders without per-user LLM calls, lifting LT30 by 6.738% in an A/B test.
A 100-task benchmark finds the frontier AI agent clears 19.1% of vision-heavy tasks where non-experts top 80%. Leaderboard scores don't transfer to deployment.
Vercel's January 2026 Vercel-Cache-Tag ship brings surrogate-key cache invalidation to every plan, moving the cache contract into application-owned tag strings.
ISSTA 2026 ablation: lightweight call topology halves code agent run variance, and forward edges in hub-heavy repos degrade results. More structure stops paying off fast.
Anthropic's MCP and Google's A2A both use JSON-RPC 2.0 but solve different problems. Here is the architectural distinction, the security gap, and when you actually need both.
A 2026 arXiv preprint open-sources an AI adoption index from chat logs and O*NET data; finance, CS, and arts top adoption. AI passes workflows but errs on specific tool calls.
Vercel AI Gateway reportedly routes GLM 5.2 Fast through Wafer, putting Zhipu's 754B-parameter MoE coding model one config entry from any Vercel app. Routing is unverified.
OpenAI is leaning toward a 2027 IPO while Anthropic's S-1 is already in SEC review, handing Anthropic the role of setting the first public frontier-model valuation template.
Vercel's tag-based cache invalidation shifts cost from each purge call to per-response edge metadata, forcing teams to design a low-cardinality tag taxonomy up front.