hardmaru ✓

@hardmaru · Minato-ku, Tokyo · joined 10 Nov 2014

Co-Founder and CEO @SakanaAILabs 🎏

436 016Followers

1 900Following

25 915Posts total

1.1MViews on collected posts

Latest posts

hardmaru ✓@hardmaru20 Sep

日経新聞による、グーグルディープマインド東京を率いる全炳河（@heiga_zen）氏の素晴らしい特集記事。 2018年、Heigaさんと2人でGoogle Brain東京チームを立ち上げた日々を懐かしく思います。当時は時差の厳しい深夜の会議をこなしながら、日本のAI研究の存在感を示すために必死でした。現在、彼が 15.9K views · 150 likes · 10 reposts · 53 replies Open on X →

hardmaru ✓@hardmaru17 Sep

Schmidhuber was building recursive self-improving systems back in 1987. His new post covers four decades of RSI, from meta-evolution and self-modifying policies to the Gödel Machine and modern LLM agents. https://t.co/gdComv0GoJ Reading this in 2026, the "pace the frontier" 68.8K views · 804 likes · 120 reposts · 43 replies Open on X →

hardmaru ✓@hardmaru16 Sep

Today everyone is talking about Recursive Self-Improvement (RSI). In 1987, when compute was 100,000,000 x more expensive, I published the 1st concrete RSI algorithms. Now compute is cheap, and RSI is driving the future of both software and physical AI. See: RSI since 1987 https:/

538K views · 1.3K likes · 193 reposts · 48 replies Open on X →

hardmaru ✓@hardmaru16 Sep

This year at Sakana AI, we built and shipped more products than I would have believed possible: Sakana Chat, Namazu, Sakana Translate, Sakana Marlin, Fugu, Fugu Cyber, and Fugu Max. We created a Product Team from zero and proved that a research lab born in Tokyo can ship https://

31.9K views · 118 likes · 12 reposts · 19 replies Open on X →

hardmaru ✓@hardmaru12 Sep

Virtual fruit fly is our generation’s Tamagotchi 🪰🧠 19.6K views · 292 likes · 13 reposts · 23 replies Open on X →

hardmaru ✓@hardmaru11 Sep

Introducing Fugu Max and Fugu Ultra v2: Orchestrating the Pareto Frontier The AI industry has spent a decade optimizing along a single axis: build bigger, more expensive models. But intelligence has never been a monolith. It is a collective, distributed system. Humanity itself h

74K views · 269 likes · 33 reposts · 32 replies Open on X →

hardmaru ✓@hardmaru11 Sep

Introducing Fugu Max and Fugu Ultra v2: the next evolution of Sakana Fugu’s multi-agent orchestration system. Try: https://t.co/aDEFyySowk Blog: https://t.co/qj3JWgNPiC The frontier that actually matters is the Pareto frontier: capability on one axis, cost on the other. But the

GIF

310.8K views · 1K likes · 181 reposts · 78 replies Open on X →

hardmaru ✓@hardmaru30 Aug

@hardmaru Except that SIs in Japan produce much lower quality software than in-house teams. They have price points to hit after all. And when the bulk of software in Japan is SI slop, adding AI slop doesn’t help. 1.7K views · 5 likes · 0 reposts · 0 replies Open on X →

hardmaru ✓@hardmaru30 Aug

@hardmaru It’s also pulled global development culture way more towards Japanese SWE norms of hyper optimized specs and a preference for hardware-specific firmware. 2.4K views · 5 likes · 0 reposts · 1 replies Open on X →

hardmaru ✓@hardmaru30 Aug

@hardmaru It is already happening; starting a business based on AI-powered contract development seems to be trending among some students. However, in urban areas, the concept is already becoming stale. 1.1K views · 4 likes · 0 reposts · 0 replies Open on X →

hardmaru ✓@hardmaru30 Aug

Silicon Valley dismissed Japan’s System Integration (SI) culture as an unscalable consultant trap. Writing the system is no longer the scarce work. Integrating it is. In the post-AI world, everyone becomes an AI-powered Japanese SIer. 69.6K views · 273 likes · 29 reposts · 38 replies Open on X →

Against accounts of the same size

11 posts from the last 90 days, next to the 100K–1M follower range. shown widely, but few of those viewers react.

Median views31 885this account5 983median for 100K–1M

Reach, %7.31%this account1.49%median for 100K–1M

Engagement, %0.46%this account0.93%median for 100K–1M

Metric	This account	Median for 100K–1M	Ratio
Median views per post	31 885	5 983	5.33×
Reach (views ÷ followers)	7.31%	1.49%	4.91×
Engagement rate	0.46%	0.93%	0.49×

Others in this range → Compare with another account → How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

69.6K30 Aug

1.1K

2.4K

1.7K

310.8K11 Sep

74K

19.6K12 Sep

31.9K16 Sep

538K

68.8K17 Sep

15.9K20 Sep

Last 11 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.51%30 Aug

0.37%

0.25%

0.29%

0.44%11 Sep

0.46%

1.68%12 Sep

0.48%16 Sep

0.30%

1.43%17 Sep

1.34%20 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for 100K–1M accounts is 0.93%.

What the audience does

Likes60.1%4 301 in total

Reposts8.3%591 in total

Replies4.7%335 in total

Quotes2.0%143 in total

Bookmarks25.0%1 786 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Latest posts

Against accounts of the same size

Growth & engagement

Views per post

Engagement rate per post

What the audience does

Similar accounts