You know someone is genuine when they can be honest about their products. There are not many people who can take that courage, but those with this trait are highly respectful. One of them is Sam Altman, the CEO behind OpenAI. He has opened up about their company’s artificial intelligence Model GPT-4 and expressed his dissatisfaction with the model.
In an interview with Lex Fridman released on Monday, Altman discussed different things regarding AI. One of the interesting moments was when the CEO opened up and said that their company’s artificial intelligence Model GPT-4 actually “kinda sucks”, and Altman hopes that the upcoming GPT-5 will truly live up to everyone’s expectations.
He added:
Looking forward over the next few years, we should realize that the tools we have now are going to be We are very backward, and this is what drives us to continue to make progress and create a better future.”
Don’t get me wrong, I don’t want to belittle the achievements of GPT-4, nor am I trying to exaggerate, Because we are on an exponential development curve, very soon we will look at GPT-3 the same way we do now. Looking at GPT-4.”
Why Sam Altman said that GPT-4 sucks?
This statement might be made to create hype for GPT-5, the upcoming model. Despite that, agreeing that the GPT-4 is not up to the mark is a commendable thing. For the most part, we have seen improvements from the GPT-3 model. However, were they drastic? Not really. Altman understands it and hence believes it can be improved to achieve greater heights.
Also, a person in the industry has seen much more than an average person. He knows what the AI is truly capable of, and that compared to GPT-4 is probably nothing. Hence, this statement makes sense, especially from his perspective. For an average consumer, even the GPT-4 is highly beneficial and can be used in any industry.