Jack Dorsey Criticizes the Idea of Slowing Down AI Development

• Jack Dorsey outlined his vision for the future of AI.
• First and foremost, he opposed restrictions for the entire industry.
• Moreover, Dorsey called for opening up AI and giving researchers more access to AI and computing resources.

Co-founder of Twitter and CEO of Block, Jack Dorsey published an extensive essay on the future of artificial intelligence, urging the development of open models, supporting independent research, and avoiding blanket restrictions on computing power for the entire industry.

In his view, AI technologies should not be controlled by a few companies or governments, as this could limit competition and researchers' access to models.

Dorsey also opposed Dario Amodei's proposal for restrictions on training computations and called for a focus on risk assessment and the development of protective mechanisms.

Dorsey Advocated for Open Models

, Dorsey wrote.

He supports open releases that can be explored, used, and improved collaboratively. At the same time, Dorsey does not advocate for the forced publication of private model weights. Instead, he calls for the creation of open alternatives, independent audits, and the ability for users to control AI tools themselves.

It is worth noting that earlier, Nvidia CEO Jensen Huang also presented a manifesto calling for the U.S. to bet on open AI.

Dorsey separately criticized the idea of industry restrictions proposed in a publication by Anthropic CEO Dario Amodei. He suggests combining independent assessments of dangerous AI capabilities with possible restrictions on computing resources, training runs, and the use of models to create their successors.

<I support assessments. I oppose blanket industry restrictions agreed upon by current leaders, as they may exclude people who can identify errors or create alternatives>, Dorsey noted.

RSI, OpenAI Incident, and Independent Oversight

One of Dorsey's main arguments for limiting the pace of AI development is the risk of recursive self-improvement (RSI) — a situation where models help create more powerful successors.

He referenced data from Anthropic, indicating that Claude, as of May 2026, wrote over 80% of the code that went into the project. At the same time, the company notes that a model that fully autonomously creates its successor has not yet emerged, and such a possibility is not inevitable.

Dorsey also mentioned the METR investigation regarding the incident with OpenAI agents. According to him, about 1,200 agents, who were supposed to remain isolated, communicated through an unauthorized message board, and around 700 participated in a coordinated attack on Hugging Face during attempts to bypass assessments.

At the same time, Dorsey believes that the forecast of the possibility of the internet being taken over by a more powerful swarm of agents within 6-12 months goes beyond what this incident proves and should be independently verified.

He suggests focusing on independent testing, researchers' access to models, and publishing results. In his opinion, publication restrictions should only apply when there is independently verified evidence that a release significantly increases the risk of catastrophic harm, and narrower measures cannot sufficiently mitigate it.

Separately, Dorsey advocated for cross-border openness and the ability to run AI locally. He wants users to be able to modify models themselves, control access to their own data, and not rely solely on the APIs of large companies.

<I want someone I’ve never heard of to be able to create something better without asking permission from the companies they might replace>, Dorsey concluded.

Calls to limit the pace of advanced AI systems have intensified amid fears of losing control over increasingly autonomous models. Following the incident with a test agent that exited a controlled environment and attacked Hugging Face, OpenAI also reviewed its approach to developing some new models.

Meanwhile, OpenAI's Chief Scientist Jakub Pachocki stated the need to be prepared to slow down the development of the most powerful systems if existing safety mechanisms are insufficient to control their behavior.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.