
Open I Chief Executive Officer (CEO) Sam Altman India is commonly called an incredible market for artificial intelligence (AI) on Wednesday. He said that Indian consumers of Openi have increased threefold in the past year. Ultman, who is in India, spoke to HT editor -in -chief R skomer Chat GPT’s future plans for India, Artificial General Intelligence and De Percycles. Edited Excerpts:

I am sure you are watching the announcements that India has made in its AI program. You were here a while ago and you made these comments – how India was better than not trying to do your frontier model – which became controversial. Has your theory changed? And do you think the Indian AI plan is on the right path?
It was in a different context. It was a different time when the frontier was extremely expensive to model. And you know, now, I think the world is a very different example. I think you can pay them at low costs and maybe an incredible job. India is generally an incredible market for AI, for us. This is our second largest market after the United States. Last year, consumers here have increased threefold. Innovation that is happening, what are people making [in India]It’s really incredible. We are excited to do a lot here, and I think this (Indian AI program) is a great project. And India will develop a great model.
What are your plans in India? Because there is there, while everyone sees the next end of the AI, it is a huge head. For example, in partnership with the Soft Bank, what you are now doing in the US is creating a huge infrastructure. Do you plan to bring some of this infrastructure to India?
We have nothing to announce today, but we are working hard, and we hope that we will share some interesting soon.
It was at the end of 2022 when you announced Chat GPT, and at the end of the week, you announced de -prescription. It seems that the pace of change is quite surprising. Micro processors have a peacock law. Is there any law related to the speed of change?
De -Percyching is something that has felt the most, such as a chat jeep, in the sense of how people are reacting. I was watching and reading online last night – I was very busy for the past two days, so I didn’t get to read the reviews – and people feel like they have any magical experience, Like them. When Chat GPT launched for the first time. So, in my opinion, this move is affecting agents from chat boats that we dreamed of at night, and it is great to see that people have another moment. The law of peacock, you know, every 18 months (every 18 months the processing power of the chips), and it has changed the world. But if you look at the cost curved letters for AI, we are able to reduce the cost of a given level of 10x (ten times) every 12 months (ten times), which is more powerful than peacock law Is If you both mix them both for more than a decade, this is something different. So although it is true that the best price of Frontier models is this steep, upward, clear [curve]Intelligence unit cost reduction rate is only incredible. And I think the world has not yet made it internal.
When was your first answer when the Chinese model, deep search news came out? At least the headline was that he managed to train his model at a very low cost, though it came out later that it was not really.
I had high doubts about the cost. It was so, some Zero is missing. But, yes, it’s a good model, and we will need to make a better model, which we will do.
AI appears to be extremely deep and capital of the most infrastructure. Is this the matter? Does this mean that there are very few players who can really work on this scale?
As we have spoken before, it’s changing. To me, the most interesting development of the previous year is how we found how to create a very powerful small model. So the Frontier will be widely expensive and it will require a lot of infrastructure, and that is why we are doing this star gate project. But, you know, we will also get a GPT4 level model on the phone at some point. So I think you can see it in any direction.
One of the challenges of where you are, and who you are, is that your company was the first company that talked about artificial intelligence and captured public imagination. When you are the first company, you have a responsibility, not only for your company, but also for the industry and how the entire industry interfaces with society. And there, there are many problems that are farming …
We have a character as I think, you are on Frontier… We have a character as a teacher, and this role is like telling society what you think is coming and you think it What is going to happen. It will not always be fine, but it is not dependent on us or any other company, to say, okay, look at this change, what society has to do.
It is up to us to say, here we see the change, here are some ideas, here are our recommendations. But society has to decide how we think how we will reduce the economic effects, how we will divide the benefits widely, how we are going to deal with these challenges. So we are a sound, an important voice in it. And I also do not mean that we do not have the responsibility of the technology we have developed. Of course we do, but this has led to a dialogue between all stakeholders.
If you look at the Indian IT industry, they have actually performed well in taking the item that other people have made very smart models in the upper part of it, and instead of making themselves models. Provided services in the upper part. Do you think they should do with AI? Or do you think, they should work more?
I think India should go for a full -fledged view of the stack …
… Which will require a lot of capital.
Well, it’s not a cheap plan, but I think it’s worth it.
You have more than 300 million users …
More …
… Okay, and what did you learn in what you are using for Chat GPT?
Can I show you something? Because this is a really meaningful thing. I was just looking at X (turning the computer to display the screen). So this guy, we are not really friends, but I know him a little. Deep Research launched a few days ago, and her daughter has a very unusual form of cancer, and she stops such a job, in my opinion, and or maybe her job is changing, And she’s working hard. He has gathered a large private research team [to understand her disease]. He has collected all this money, and deep research is giving him better answers to the private research team he has hired. And seeing such things as really meaningful for us.
Do you expect President Trump to take further steps to protect US leadership in AI? Do you see that is happening? Or, to make the question differently, will any national game be played in AI?
Of course there is. But our mission, which we take very seriously, is for AGI (Artificial general intelligence) To benefit all humanity. I think this is one of the rare things that goes beyond national borders. Ai wheels and fire, industrial revolution, is like the agricultural revolution and it is not a country. This belongs to everyone. I think AI is one of these things. This is like the next step. And they do not belong to the nations.
You first talked about artificial general intelligence a few years ago. Have we gone close to it?
Yes, when I think about the models are now worth what they can do a few years ago. I think we’re indefinitely close …
Are we more brave now with our Felicafs?
Where we have moved a few years ago … I think we have made progress in the safety and strengthening of the model two years ago. You know, look at the ability to comply with the current model’s clear rate, or a set of policies, we are in a better state than two years ago. This does not mean that we do not have to solve the thoughts of Superintendees (AI or a theoretical construction of intelligence or more ideological construction than human intelligence). Of course we do, but we are there at an excellent speed.
Have you seen Lancet Paper on Swedish breast cancer studies that came out yesterday? They used an AI model called Transpare, which I do not know if you are familiar with it, and they discovered that the correct diagnosis has increased by 29 %, without any wrong positive…
This is terrific. I was thinking the other day, you know, how much better is AI to drive? How much do AI have to be better than a human doctor before allowing the diagnosis? This is clearly improved. Self -run cars should be much safer than human drivers to accept the world. But, how many of these studies do we need before we want the AI doctor?
Although I just think that when it comes to diagnosis, the bar will be less than the cars …
I think for cars, maybe the subject, you want it to be safer 100 times. IT of diagnosis, it should be Very little