World models, artificial general intelligence and the hard problems of life–mind continuity: toward a unified understanding of natural and artificial intelligence https://t.co/pl5XJkB73x
Who's Who of AI
David Ha
Why they matter
Tracked through public AI activity and peer connections inside the directory.
- AI signals
- 2 past 30d
- Sources
- 2 distinct domains
- Discussions
- 0 past 30d
- Latest signal
- 6d ago
What they're sharing
royalsocietypublishing.org
Sakana AI sakana.ai
Sakana Fugu Technical Report arxiv.org
Sakana AI sakana.ai
Fugu Ultra - API Pricing & Providers openrouter.ai
Sakana Chat chat.sakana.ai
jst.go.jp
Sakana AI sakana.ai
Sakana AI sakana.ai
Sakana AI sakana.ai
Sakana AI sakana.ai
Sakana AI sakana.ai
Articles & links
Sakana Fugu Technical Report https://t.co/6e6WuA8FVB Release Notes: https://t.co/7xWGpOicFN https://t.co/g2yaZvex35
RT @SakanaAILabs: Sakana Fugu Technical Report https://t.co/BRGepSmyI5 🐡 https://t.co/7d3webOkmH
AI Weekly's analysis
→
- Fugu-Ultra reportedly scores 73.7% on SWE-Bench Pro and 82.1% on Terminal Bench 2.1, beating Claude-Opus-4.8, Gemini-3.1-Pro, and GPT-5.5 baselines.
- Fugu is itself a language model that learns to orchestrate Gemini-3.1-Pro, Claude-Opus-4.8, and GPT-5.5 as expert workers per query.
- Training combines supervised fine-tuning with sep-CMA-ES evolutionary optimization for Fugu, and GRPO reinforcement learning for Fugu-Ultra.
Read full analysis →
@SakanaAILabs Introducing Fugu Max and Fugu Ultra v2: Orchestrating the Pareto Frontier https://t.co/OzNZkTJKyu 🐡 https://t.co/KkMjRuPw9S
AI Weekly's analysis
→
- Fugu Max prices at $2 per million input tokens and $6 per million output tokens, which Sakana claims runs 40-60% below Sonnet 5, GPT 5.6 Terra, and Kimi K3.
- The system routes each task across a swappable pool of open-weight and specialized models, with NVIDIA's Nemotron folded in via an August 2026 collaboration.
- Fugu Ultra v2 scores 48.3 on Chartography against Opus 5's 27.3, and does so without Fable 5, Fable 5.1, or GPT-6-Astra in its agent pool.
Read full analysis →
@OpenRouter https://t.co/kxUkYpSHHF
AI Weekly's analysis
→
Read full analysis →
Here is a quick 60-second summary of vibe-coding in Sakana Chat: • An interactive math game for kids • A fully cited market research report • A Bank of Japan interest rate analysis • A vessel monitor for the Strait of Hormuz Use it for free: https://t.co/FCGApHFkjW 🐟 https://t…
AI Weekly's analysis
→
Read full analysis →
RT @SakanaAILabs: JST-CRDS（科学技術振興機構 研究開発戦略センター）のショートレポートに、Sakana AIのAI Scientistが、日本発の代表的な取り組みとして取り上げられました。 https://t.co/8z4qORaf25 同レポート…
RT @SakanaAILabs: ベースモデルに依存しないオーケストレーションに向けて ブログ：https://t.co/Dc1STD2SSB Sakana Fuguは、マルチエージェントのオーケストレーションを一つの基盤モデルとして提供するプロダクトです。一つのエンドポ…
AI Weekly's analysis
→
- Sakana AI retrained its Fugu conductor model on Gemma 4 E2B and reports comparable performance to the earlier Qwen-based version.
- Fugu is a two-layer setup: a small conductor routes work to a pool of frontier models the company calls hundreds-of-billions-scale.
- Sakana frames the result as base-model-agnostic training, groundwork for a future conductor built on a domestically-developed base model for sovereignty use cases.
Read full analysis →
I’m looking to hire a Program Manager to help manage Sakana AI’s fast growing Recursive Self-Improvement (RSI) Lab 🚀 RSI Lab (English): https://t.co/Sz46xHIfZi RSI Lab (日本語): https://t.co/gCBtwg5imq Job Description: https://t.co/DoWrN7jSl9 https://t.co/VlsBVCQOHI
AI Weekly's analysis
→
- Sakana AI has launched a dedicated Recursive Self-Improvement Lab in Tokyo, framed around redesigning the AI development process with AI itself.
- The lab points to prior work such as the Darwin Gödel Machine, reported to drive a 30 percentage point absolute improvement on SWE-bench.
- Sakana positions RSI as reachable on modest compute and aligned with Japan's sovereign AI strategy, rather than hyperscaler-style scaling.
Read full analysis →
Program Manager (RSI Lab) https://t.co/DoWrN7jSl9
AI Weekly's analysis
→
- Sakana AI is hiring a bilingual program manager in Tokyo for its Recursive Self-Improvement Lab, positioned as a dedicated research group.
- The lab says its Darwin Gödel Machine more than doubled baseline SWE-bench software-engineering performance, a 30 percentage point absolute improvement.
- Sakana says it is building the most sample-efficient self-improvement engine, not the most compute-hungry, backed by Japan's sovereign AI push.
Read full analysis →
I’m looking to hire a Program Manager to help manage Sakana AI’s fast growing Recursive Self-Improvement (RSI) Lab 🚀 RSI Lab (English): https://t.co/Sz46xHIfZi RSI Lab (日本語): https://t.co/gCBtwg5imq Job Description: https://t.co/DoWrN7jSl9 https://t.co/VlsBVCQOHI
AI Weekly's analysis
→
- Sakana AI has formally established a Tokyo-based research group dedicated to recursive self-improvement using foundation models.
- The lab points to its Darwin Gödel Machine, which reportedly more than doubled its baseline software-engineering performance on SWE-bench.
- Sakana frames the bet as sample efficiency rather than raw compute, and is hiring Frontier Research Scientists and Advanced Core Engineers.
Read full analysis →
RT @SakanaAILabs: Join the team behind Sakana Chat 🐟, Marlin 🐬, and Fugu 🐡 https://t.co/jzoepOf44n
AI Weekly's analysis
→
- Sakana AI has six open product roles based in Tokyo, spanning applied research, engineering, product management, sales, marketing, and design.
- All six postings sit behind three named products: Sakana Chat, Sakana Marlin, and Sakana Fugu.
- Marlin, launched in June as Sakana's first commercial product, is a research agent that runs for up to eight hours to produce 100-page reports.
Read full analysis →
Are you David Ha? Show it.
Add the Who’s Who of AI badge to your site or bio. It links back to this profile.
Markdown: [](https://aiweekly.co/whos-who/person/david-ha)