Bienvenue sur mon blog
Reflexions sur la technologie, le developpement et l'innovation
Articles récents
Microsoft's MAI-Thinking-1 goes toe to toe with Claude Sonnet 4.6
Microsoft dropped its first reasoning model at Build 2026 Day 2, and the benchmarks are worth paying attention to. MAI-Thinking-1, a 35-billion active parameter Mixture of Experts model (roughly 1 trillion total parameters), claims human preference parity with Anthropic’s Claude Sonnet 4.6 in blind evaluations. It also matches Claude Opus 4.6 on SWE-Bench Pro. That is a lot of firepower for a model that costs less to run than anything in Opus or Sonnet’s weight class. The key claims from the Mic
Lire l'articleMicrosoft builds its own coding AI and cuts the OpenAI cord
Microsoft Build 2026 opened today in San Francisco and the headline was not Windows. Satya Nadella unveiled Project Polaris, a homegrown AI coding model that will replace GPT-4 Turbo as the default brain inside GitHub Copilot starting August 2026. Automatic migration for all Copilot subscribers, with a three-month fallback window for teams not ready to switch. This is not a minor model swap. Polaris runs on Microsoft’s custom Maia 200 AI accelerators inside Azure, which the company says cuts per
Lire l'articleFirst confirmed LLM agent cyberattack exfiltrated AWS database in under an hour
Sysdig documented something the security industry has been warning about but had not yet seen in the wild: a large language model agent autonomously performing post-exploitation in a real attack. The target was an AWS-hosted PostgreSQL database. The entire chain, from initial access to exfiltration, took a little over an hour. The attack happened on May 10, 2026. The entry point was CVE-2026-39987, a pre-authentication remote code execution vulnerability in Marimo notebooks. An unauthenticated a
Lire l'articleNightmare Eclipse released six Windows zero-days in six weeks before GitHub and GitLab pulled the plug
A pseudonymous security researcher operating under the name Nightmare Eclipse spent six weeks systematically dropping working Windows zero-day exploits on GitHub. Six exploits. Six weeks. All targeting different parts of the Windows security architecture. Two code-hosting platforms eventually banned the account. The exploits are still out there. The campaign started April 3 with BlueHammer, a Windows Defender local privilege escalation flaw. The PoC gave attackers a SYSTEM shell via a TOCTOU and
Lire l'articleAI found 10000 critical bugs and the patches still are not keeping up
Anthropic published the first progress report on Project Glasswing on May 22. The numbers are staggering. In roughly 30 days, Claude Mythos Preview and about 50 partner organizations flagged 23,019 potential vulnerabilities in open-source software. Of those, 6,202 were classified as high or critical severity across more than 1,000 projects. Independent validation confirmed 1,726 as true positives. 1,094 earned a high or critical rating. Only 97 have been patched. That gap between discovery and r
Lire l'articleAnthropic ships Opus 4.8 and a $65 billion round on the same day
Anthropic released Claude Opus 4.8 and closed a $65 billion Series H at a $965 billion valuation within hours of each other. The timing was not accidental. The new model landed 41 days after Opus 4.7. That is unusually fast for an Anthropic flagship. The most recent Sonnet and Haiku models are three and seven months old respectively. The accelerated cycle was likely driven by the lukewarm reception Opus 4.7 got from developers, combined with competitive pressure from OpenAI’s Codex and Google’s
Lire l'article