OpenAI CEO Sam Altman visits Taiwan｜Taiwan News

Yonglin Foundation held an AI Forum themed “Redefining Tomorrow: The AI Revolution." Among the speakers was Sam Altman, the CEO of OpenAI. During his speech, he emphasized his optimism about the future of OpenAI and promised to continue optimizing ChatGPT and solve the risks that AI poses. FTV reporter Stephany Yang has the details.

This is Sam Altman’s third visit to Taiwan. Speaking at an AI forum in Taipei, he said that despite public doubts about AI, he is optimistic about the future of OpenAI. He sees great potential in employing OpenAI in the fields of education and health care.

Sam Altman
OpenAI CEO
We have been calling for regulation but only for the most powerful systems. Let the small models go. Let open source go. Models that are like 10,000 times the power of GPT-4. Models that are as smart as human civilization. Those probably deserve some regulation. It seemed to me like the least controversial thing that I was going to say this year. I got and still am getting bashed for it by the tech industry. There is this reflexive anti-regulation thing but regulation has not been a pure good. But it has been good in a lot of ways. Some areas I am particularly excited about are about are education, health care, and personal productivity tools like what we see people do using those coding tools that has been amazing.

When asked about AI’s risks like misinformation and reinforcement of biases, Altman said it was important to set boundaries and input human feedback.

Sam Altman
OpenAI CEO
It is certainly true that if you just train a model on the raw Internet you are going to get something that most reasonable people will call biased in different ways. But with techniques that we and others have developed like reinforcement learning from human feedback and a whole bunch of other things, you can surprisingly steer these models. And so, you can figure out how you want OpenAI to make some decisions but you can change it a lot. This idea of empowering individuals, but setting the edges, and also the sort of standard behavior well I think is very important.

Speaking on the ways to make AI more sustainable, Altman said that fusion or solar nuclear-powered green energy may be able to provide a safe, clean, and cheap source of power at scale.

Sam Altman
OpenAI CEO
Fusion and if that doesn’t work, solar. The energy needs as we look out are enormous. I think the cost of intelligence over time should trend towards the cost of energy. There is going to be a huge demand for intelligence and, thus, a huge demand for energy. If we can create abundant, cheap, clean, safe energy. And again, I think we are going to have this very soon, fusion at a massive scale. Like the scale that you need for everybody on earth to get to enjoy the fruits of AGI. That is our plan.

Experts speaking at the forum said that while there are concerns surrounding AI, they hope to work together to solve its risks.

For more Taiwan news, tune in:
Sun to Fri at 9:30 pm on Channel 152
Tue to Sat at 1 am on Channel 53

#台灣新聞 #TaiwanNews #民視新聞 #FTV新聞 #Taiwan

[masteroogway]

hello where can i find the full video
thanks i am a very desperate student