Archive
Every issue of the AlphaSignal newsletter, newest first.
649 issues total
June 2026
28 issues- Jun 30 Meta releases brain-to-text decoder, Cursor ships iOS app, MegaTrain runs 120B models on one GPU, DeepSeek boosts LLM throughput 400%
- Jun 29 OpenAI previews GPT-5.6 family, Nous Research beats Claude on benchmarks, LLMs reuse 11 words in 88% of stories, Codebase-memory indexes Linux in 3min
- Jun 28 The autonomous coding dream just got cheaper. GLM-5.2 is forcing a rethink on closed AI.
- Jun 27 Voice is the new keyboard. See how developers are prompting AI agents 4x faster with Wispr Flow.
- Jun 26 Qwen ships open-source world model beating GPT, Meta releases Autodata agent, Zyphra finds LLMs lose ability to learn, OpenCode goes open source
- Jun 25 OpenAI reveals custom Broadcom chip, Mistral ships OCR 4 with 72% win rate, Krea 2 shrinks to 12GB, Hermes Agent learns reusable skills
- Jun 24 Sakana ships export-control bypass model, Claude joins Slack as coder, GLM-5.2 beats Opus at half cost, Bengio warns on alignment
- Jun 23 OpenAI launches GPT-5.5-Cyber, Cursor adds SpaceX-trained model, GLM 5.2 tops open-source coding, looped transformer hits 94% on Sudoku
- Jun 22 Berkeley releases 2.2M laws dataset, Stanford ships cited research tool, critic-free training beats GRPO by 18%, Nous ships Hermes Agent v0.17.0
- Jun 21 Stop tweaking prompts by hand. Here's how agents can now optimize their own skills automatically
- Jun 19 Unsloth shrinks GLM-5.2 by 84%, OpenAI trains GPT-5.5 with 600 doctors, Berkeley turns videos into robot training data
- Jun 19 Midjourney reveals hardware, Cursor ships cloud agents, OpenAI Codex records workflows, Anthropic adds Artifacts to Claude Code
- Jun 18 Anthropic launches Claude Design, xAI ships image-to-video model, OpenCut hits 55K stars, Zhipu ships GLM-5.2 with 1M context
- Jun 17 Z.ai ships GLM-5.2 with 1M context, Cursor launches Origin git platform, Hermes agent buys via Stripe, Gemma 4 runs on 12GB VRAM
- Jun 16 Sarvam raises $234M, MiniMax opens weights, Kimi K2.7 beats GPT-5 on math, Ng ships free local AI agent, transformer burned into custom chip
- Jun 15 Anthropic kills two Claude models, Kimi open-sources coder, LMCache boosts LLM inference 10x, Google maps AGI paths
- Jun 14 The top AI engineers aren't prompting anymore, they're designing loops. Here's how to build them without burning your A
- Jun 12 OpenAI banks Codex rate limits, Kimi open-sources coding model, 810K-star agent prompt library, Goodfire audits model training
- Jun 11 Cohere ships 30B coding model, Google translates 70+ languages live, Claude builds Minecraft clone for $30, Microsoft opens free ML course
- Jun 10 Cursor ships Claude at 72.9%, Anthropic adds scheduled agents, Cohere releases 30B open-source coder, Perplexity study cuts task time 87%
- Jun 9 Anthropic ships Claude 4, Cognition drops 13/100 coding benchmark, Google upgrades NotebookLM, Kimi launches 300-worker agent
- Jun 8 Claude Opus 4.7 matches chemistry software, Cursor ships Design Mode, Google releases multi-agent RAG, Liquid AI ships Japanese audio model
- Jun 7 Async agents generate 741% more code. Releases? Up just 20%. Here's why the math doesn't add up.
- Jun 5 Claude hits 52x code speedup, OpenAI ships smarter memory, Codex builds iOS apps in browser, NVIDIA drops Nemotron Ultra
- Jun 4 Robot planner runs 19x faster, Unsloth compresses Gemma 4 12B, NVIDIA ships Cosmos 3 image model, 8B TTS hits 110ms latency
- Jun 3 Anthropic files IPO paperwork, OpenAI Codex builds websites, Nous ships Hermes Desktop, Datalab releases 650M OCR model Surya
- Jun 2 JetBrains releases Mellum2 12B, LiquidAI opens 8B MoE for fine-tuning, 800K model beats frontier LLMs on Sudoku, robot model trains on 27K hours
- Jun 1 OpenAI launches biodefense platform, xAI releases coding model at $1/M tokens, Alibaba ships one model for 11 robots, Cursor adds auto-review mode