• AI Periscope
  • Posts
  • The world is not ready for ChatGPT-5 says OpenAI employee

The world is not ready for ChatGPT-5 says OpenAI employee

And: Microsoft prez - "we better have a way to slow down or turn off AI" | AI researchers suggest brakes on compute hardware

Author

AI Periscope
February 22, 2024

ChatGPT-5

Summary - OpenAI is concerned about society's readiness for advanced AI systems like ChatGPT-5. It has slowed the release of certain AI technologies, allowing the public to gradually absorb the new capabilities and consider the impacts. This cautious approach has led to a divided public opinion, with excitement about AI's potential and apprehension about job displacement, ethical challenges, safety, and the need for effective regulation. Addressing these issues requires collaborative efforts, which may be slow to come in an already-divisive society.

POV - The release of OpenAI’s Sora last week was met with excitement and fanfare… at first. It didn’t take long, however, for concerns of a dystopian world to quell that excitement - perhaps not totally unfounded, either. Sam Altman’s firing and re-hiring is still curious to me. It was supposedly over a letter employees wrote to the OpenAI board warning of some incredible breakthroughs. What are they not telling us?

Turn-off AI?

Summary - In a recent interview, Microsoft President Brad Smith discussed artificial intelligence and the need for regulation and limits. He compared AI to the existential threat in the movie, The Terminator, and emphasized the importance of having a way to slow down or turn off AI, especially in critical infrastructure systems. Smith also called for an emergency brake on AI, similar to the brakes on buses and trains, to prevent potential harm to humanity. However, there is a debate on who should regulate AI, with some questioning the role of organizations like Microsoft.

Key points:

  • Smith thinks AI is as important of an invention as the printing press.

  • On jobs displacement, Smith compares our need to adapt to AI like that of our adaptation to PCs and mobile phones 40+ years ago.

  • He emphasizes the need for regulation and safety measures.

  • Smith suggests the implementation of emergency brakes for AI to address concerns about its potential to harm humanity.

  • The question of who should regulate and control AI remains, with the need for collaboration between organizations, experts, and governing bodies.

POV - As much as there is excitement for AI, concerns for the technology is everywhere you look as well. Regulation is slow, which is not surprising given this uncharted territory. Makers of AI technologies are joining forces and imposing some self-governance to some extent, but will it be enough?

Hardware off switch

Summary - AI researchers have released a 104-page document discussing the risks and potential regulations surrounding artificial intelligence. They propose putting regulations on the compute hardware itself as a potential solution. Some highlights from the document include the risks of increased AI hardware monitoring on personal privacy, the leakage of sensitive information to policy makers, negative economic impacts on labor markets, the concentration of power in AI compute regulation, and the use of low-compute specialized models with dangerous capabilities. The document suggests potential solutions such as a global registry of AI chips and the use of "kill switches" to deactivate AI hardware being used for malicious purposes.

POV - Pretty bleak outlook for a Thursday edition, wouldn’t you say? Fear not. I don’t think it’s all doom and gloom. I am still bullish on humans, and believe that we will find the right solution. That has been the case throughout history. It won’t be a straight line - we’ll zig some, and zag some. Even though there is typically high turbulence in the periods immediately following big paradigm-shifts, humans have always found progress in the longer term. Do you agree? Are you more optimistic or pessimistic about AI?

Don’t forget to check out other AI headlines in The Ocean.

The Ocean is designed to help you stay informed with the wide array of AI headlines, giving you just the snippet of news that you need if you are short on time, but also gives you one-click ease if you want to dive deeper.