hardmaru
hardmaru.bsky.social
hardmaru
@hardmaru.bsky.social
Co-Founder & CEO, Sakana AI 🎏 → @sakanaai.bsky.social

https://sakana.ai/careers
Reposted by hardmaru
RePo moves us toward models that intelligently curate their own working memory rather than passively accepting input order.

Read the full breakdown on our website:
pub.sakana.ai/repo/

Paper: arxiv.org/abs/2512.14391
RePo: Language Models with Context Re-Positioning
In-context learning is fundamental to modern Large Language Models (LLMs); however, prevailing architectures impose a rigid and fixed contextual structure by assigning linear or constant positional in...
arxiv.org
January 19, 2026 at 12:40 AM
Reposted by hardmaru
Introducing RePo: Language Models with Context Re-Positioning

Standard LLMs force a rigid linear structure on context, treating physical proximity as relevance. Cognitive Load Theory suggests this is inefficient—models waste capacity managing noise instead of reasoning.

arxiv.org/abs/2512.14391
January 19, 2026 at 12:39 AM
Reposted by hardmaru
2026 is just getting started 🚀✨

We are hiring. Join our team in Tokyo!

sakana.ai/careers
January 14, 2026 at 1:23 PM
Reposted by hardmaru
AI導入は「雇用不安小さい正社員制度が強みに」

日経ビジネスにて、Sakana AI CEO @hardmaru.bsky.social
のインタビューが公開されました。企業へのAI実装が本格化する2026年における現状と課題、そして日本企業の組織文化がAI導入にとってポジティブに働く可能性について語りました。

business.nikkei.com/atcl/gen/19/...

【記事のハイライト】🧵
サカナAIのデビッド・ハCEO、AI導入は「雇用不安小さい正社員制度が強みに」
AIファーストを掲げる企業が増加する中、経営者はAIのリスクを正しく理解し、適切に導入を進める必要がある。国内最大級のユニコーンで、企業向けのAIソリューション開発を行うSakana AI(サカナAI、東京・港)のデビッド・ハ最高経営責任者(CEO)に、日本企業のAI導入における課題を聞いた。
business.nikkei.com
January 13, 2026 at 8:46 AM
Reposted by hardmaru
Introducing DroPE: Extending Context by Dropping Positional Embeddings

We found embeddings like RoPE aid training but bottleneck long-sequence generalization. Our solution’s simple: treat them as a temporary training scaffold, not a permanent necessity.

arxiv.org/abs/2512.12167
pub.sakana.ai/DroPE
January 12, 2026 at 4:07 AM
One of my favorite findings: Positional embeddings are just training wheels. They help convergence but hurt long-context generalization.

We found that if you simply delete them after pretraining and recalibrate for <1% of the original budget, you unlock massive context windows. Smarter, not harder.
Introducing DroPE: Extending Context by Dropping Positional Embeddings

We found embeddings like RoPE aid training but bottleneck long-sequence generalization. Our solution’s simple: treat them as a temporary training scaffold, not a permanent necessity.

arxiv.org/abs/2512.12167
pub.sakana.ai/DroPE
January 12, 2026 at 4:12 AM
Reposted by hardmaru
We are taking our technology far beyond competitive programming to unlock a new era of AI-driven discovery.

We are hiring. Join our team in Tokyo.

sakana.ai/careers/#sof...
January 10, 2026 at 3:07 AM
We are taking our technology far beyond competitive programming to unlock a new era of AI-driven discovery.

We are hiring. Join our team in Tokyo.

sakana.ai/careers/#sof...
January 10, 2026 at 3:08 AM
When agents compete for limited resources, intelligence reorganizes around survival, not elegance.
January 9, 2026 at 9:26 AM
Survival of the fittest code!

Our paper explores LLMs driving an evolutionary arms race in Core War, where assembly programs fight each other. We task LLMs with evolving "Warriors" in a virtual machine, producing chaotic, self-modifying code dynamics.

Blog: sakana.ai/drq
Paper: pub.sakana.ai/drq/
Introducing Digital Red Queen (DRQ): Adversarial Program Evolution in Core War with LLMs.

In this work, we explore how LLMs can drive open-ended adversarial evolution of programs within the Core War environment.

Blog sakana.ai/drq
Website pub.sakana.ai/drq/
ArXiv arxiv.org/abs/2601.03335

Thread:
January 8, 2026 at 5:11 PM
Reposted by hardmaru
Introducing Digital Red Queen (DRQ): Adversarial Program Evolution in Core War with LLMs.

In this work, we explore how LLMs can drive open-ended adversarial evolution of programs within the Core War environment.

Blog sakana.ai/drq
Website pub.sakana.ai/drq/
ArXiv arxiv.org/abs/2601.03335

Thread:
January 8, 2026 at 5:00 PM
So proud of Team Sakana AI for pulling this off! We managed to get an agent to rank #1 in a difficult heuristic optimization contest. We leaned heavily into test-time inference using a mix of frontier models.

The agent spent $1,300 to autonomously discover an algorithm that beat the human baseline.
Our AI agent has achieved 1st place in a competitive optimization programming contest against over 800 human participants.

Blog: sakana.ai/ahc058

Thread:
January 5, 2026 at 4:01 PM
Reposted by hardmaru
Our AI agent has achieved 1st place in a competitive optimization programming contest against over 800 human participants.

Blog: sakana.ai/ahc058

Thread:
January 5, 2026 at 3:53 PM
Happy New Year! ⛩️
明けましておめでとうございます。

2025年、Sakana AIは事業開発組織を始動させ、研究の成果を社会実装へと繋げる確かな一歩を踏み出しました。 旧年中、私たちと共に歩んでくださった皆様に深く感謝いたします。

2026年、AIはどの産業でも本格実装の時代に突入します。私たちは、最先端のオープンエンドな研究開発を追求し続けると同時に、その成果を日本の産業へと実装し、昨年着手した実装をスケールさせていきます。 日本が世界のAIエコシステムにおいて重要な一角となるよう、業界の皆様と切磋琢磨し、新たな価値を創造していく所存です。
January 1, 2026 at 3:47 AM
Sakana AI’s office looks like this.
【Sakana AI 採用】 日本の未来を切り拓く、AIの社会実装を支える、オフィスアドミニストレーターを募集します!

急成長中のスタートアップで、快適なオフィス環境作りやメンバーのサポートなど、幅広く活躍してみませんか?

詳細・応募はこちら:
sakana.ai/careers/#off...
December 28, 2025 at 1:43 AM
Software Engineering as a profession will continue to fundamentally change in 2026.

Humans will need to learn to co-adapt to this evolving “alien technology” which comes with no real manual, and figure out how to operate it.

What a time to be alive ✨

twitter.com/karpathy/sta...
December 27, 2025 at 1:09 AM
Reposted by hardmaru
Merry Christmas! 🎄

Sakana AIでは、事業開発に関心がある方向けの「カジュアル面談窓口」をオープンしました!金融・防衛・インテリジェンス領域で、私たちがどのような開発に挑んでいるのか。中の人が直接お話しします。

募集職種: エンジニア、Project Manager、Product Manager 内容: 事業戦略、開発の裏側、チームの雰囲気など

最先端のAI開発を社会実装するプロセスに興味がある方、ぜひお気軽にご応募ください!

👉 応募フォームはこちら: forms.gle/sW5wz23SLSvN...

👉 募集要項: sakana.ai/careers/
December 25, 2025 at 7:47 AM
I doubt that anything resembling genuine AGI is within reach of current AI tools—Terence Tao

mathstodon.xyz/@tao/1157223...
December 22, 2025 at 7:44 AM
“iRobot Corp., the company that revolutionized robot vacuum cleaners in the early 2000s with its Roomba model, filed for bankruptcy and proposed handing over control to its main Chinese supplier.” 😥
www.bloomberg.com/news/article...
Robot Vacuum Roomba Maker Files for Bankruptcy After 35 Years
iRobot Corp., the company that revolutionized robot vacuum cleaners in the early 2000s with its Roomba model, filed for bankruptcy and proposed handing over control to its main Chinese supplier.
www.bloomberg.com
December 15, 2025 at 12:26 PM
『日経ビジネス』のインタビュー記事が公開されました。

日本の組織構造は長年の知恵の結晶であり、無理にフラット化すべきではありません。特性の異なるAIを組み合わせ、既存の蓄積を代替するのではなく、人に寄り添う「コンパニオン」であるべきです。組織の強みを活かすAIの在り方を語りました。
生成AIは「人の成長を加速させるパートナー」になれるか

Google Cloud様提供の『日経ビジネス』記事にて、Sakana AI CEO David Haが参加したインタビューが公開されました。業務に適したAIモデルの組み合わせ方や、日本企業の強みを活かすAI導入について語りました。

special.nikkeibp.co.jp/atclh/ONB/25...
December 14, 2025 at 11:19 PM
Reposted by hardmaru
Sakana AIでは、🐟 Recruiterを募集しています!

先端AIの社会実装をミッションとするApplied Teamの一員として、EngineerやProject Managerなどの採用を担っていただきます。ダイレクトソーシングを軸に、候補者一人ひとりと直接的な関係を築くことで、Sakana AIの成長を加速させる重要な役割です。

JD: sakana.ai/careers/#rec...

IT/テクノロジー業界でのリクルーティング経験を活かし、Sakana AIのチーム組成をともに担っていただける方のご応募をお待ちしております!
December 14, 2025 at 5:07 AM
“Why AGI Will Not Happen” by Tim Dettmers.

timdettmers.com/2025/12/10/w...

This essay is worth reading. Discusses diminishing returns (and risks) of scaling. The contrast between West and East: “Winner takes all” approach of building the biggest thing vs a long-term focus on practicality.
December 14, 2025 at 3:04 AM
Reposted by hardmaru
Sakana AIでは、研究成果を社会実装につなげる「Applied Team」を急速に拡大しています。

世界最先端の自律型エージェント技術を駆使して、未踏のソリューション開発に挑むApplied Research Engineerを募集中です。技術の社会実装をさらに加速させる、コアメンバーとしての参画をお待ちしています🚀

sakana.ai/careers/#app...

(正社員・学生インターン問わず歓迎です✨)
December 12, 2025 at 10:35 PM
Reposted by hardmaru
“We want to build AI models that actually understand Japanese context”

An interview with Tarin Clanuwat published by the Tokyo Government.

Read about our initiative to create AI that appreciates Japanese culture along with its rich historical contexts:

tokyoupdates.metro.tokyo.lg.jp/en/post-1649/
December 10, 2025 at 9:09 PM
Reposted by hardmaru
Where will entry-level developers gain the experience they need, as AI takes over grunt work?

At Open Source Summit Japan, Sakana AI researcher Stefania Druga suggests AI could help solve this conundrum.
thenewstack.io/ai-is-killin...
AI Is Killing Entry-Level Programming Jobs. But Could It Also Help Save Them?
Where will entry-level developers gain the experience they need, as AI takes over grunt work? At Open Source Summit Japan, an AI researcher suggests AI could help solve this conundrum.
thenewstack.io
December 10, 2025 at 2:09 PM