Strategies and dangers hidden in the request to stop AI development[Opinion]| coindesk JAPAN | Coindesk Japan

Tech moguls signed a letter last month calling for a six-month moratorium on training AI models more powerful than GPT-4.

This letter is dangerous and will provoke thoughtful citizens. The signatories argue that the pause in development will give them more time to understand and respond to the risks inherent in AI.

true aim

But if it does, it will irreparably hurt America’s economy and people.

GPT-4 and similar models promise to boost human capabilities a thousandfold and drive social change in many areas of life. The current dynamics in the industry are likely to dictate who benefits from this technology.

For example, imagine that in 1997, Microsoft and Dell issued similar “stop development” letters. Think of it as calling for a six-month moratorium on web browser development and a ban on e-commerce sites, stating that the Internet is destroying brick-and-mortar businesses and supporting terrorist funding.

The letter can be understood as an attempt to incite fear and effectively control regulation to their own advantage.

letter danger

A few glamorous leaders are trying to seize power while pretending to protect us from the dangers of AI. What technology can the world have and how leaders can use it? We are going to be the only ones that can decide what makes AI “accurate, safe, interpretable, transparent, robust, coordinated, trustworthy and faithful.”

It is wrong for a handful of billionaires to decide what is good and safe for the world. Not even well-intentioned leaders in the AI ​​industry should have such power. Absolute power corrupts absolutely.

The world is now in the middle of the next-generation AI development race.

None of the race participants stopped or slowed down research and development. Independent AI labs and international rivals are desperate to incorporate advanced AI into their systems, and there’s no stopping them. He keeps running relentlessly.

learning from blockchain

How can we ensure that everyone benefits from the 1,000-fold increase AI promises?

The only way to do that is through free and open development, including sharing features, methods and network checkpoints. For example, over the years, EleutherAI has defied false warnings that GPT-2 and GPT-3 are “too dangerous” for the public, fearlessly releasing research and models to pave the way. Globally, we need 100 to 1,000 times the scale of EleutherAI going forward (note: I have no affiliation with EleutherAI).

You shouldn’t stop walking. We should now prioritize, invest in, contribute to, and publish a truly open AI model. We must remain vigilant against those who seek to control the fate of mankind.

Eleven years ago, pioneers in the blockchain space came together to form the Bitcoin Foundation. A model of organization in the crypto industry that balances the needs and goals of private individuals with those of businesses that enable crypto to thrive.

The Bitcoin Foundation included many different types of people and many goals, but one thing in common was the desire to make Bitcoin a success without being controlled by a single group or faction. I believe AI technology needs a similar organization. Learn the lessons of the last decade or so of blockchain and combine them with a resolute commitment to independence, humanity and openness.

The moratorium on development proposed by the letter would cement the domination of AI development by the rich and powerful. Instead, let us race forward together in the open.

I would like to discuss with like-minded people, skeptics and curious people how we can achieve a better future. At freeaimovement.com, you can connect with those working to make AI free and open for all.

Mr. Peter Vessenes: Technologists and investors with deep roots in digital currency, tokenization and blockchain. He was involved in Bitcoin early on and co-founded the Bitcoin Foundation in 2012. He is also the first cryptocurrency industry insider to advise the US Government Accountability Office, US Treasury Department, and other organizations on Bitcoin.

｜Translation and editing: Akiko Yamaguchi, Takayuki Masuda
｜Image: Shutterstock
｜Original: The AI ​​’Pause’ Proposal Is Deceptive and Alarmingly Hazardous