As much as the technology of the future is called, and almost from the present, the truth is that artificial intelligence generates uncertainty. Perhaps because many experts warn about it every day, either because they anticipate that it will leave everyone without work, or even extinguish humanity. From OpenAi they are aware of it.
That is why Sam Altman’s company, after incidents that even seems to have jeopardized the integrity of some users, security is being taken very seriously. So much, that Richard Ho, the company’s hardware chief has spoken about it. And he has done so categorically, that his statements have not taken long to go around the world.
Sam Altman has reasons to worry about OpenAi
It is not usual for a member of a company to be so hard when talking about its product. It almost seems as if stones are thrown against your own roof. But the truth is that Richard Ho has played such a sensitive issue for OpenAi as the security of artificial intelligence. Was there in your comments a pull of ears for Sam Altman himself?
The expert has referred to the importance it has, and that will above all will have a precise control over the hardware. Even He has talked about the need to have a «safety switch». That is, of a kind of «red button» with which to disconnect chatgpt if the thing gets ugly, like some predicting. Something Sam Altman denied a while ago.
«It has to be integrated into the hardware,» HO says, as the Techradar specialized medium collects. The person in charge ensures that it would be a mistake to give things. «Today, much of the security work focuses on software. It is assumed that hardware is safe. It is assumed that the hardware will work properly. It is assumed that hardware can be disconnected. »
However, he is not clear, and considers that a good part of his work is to increase the protocols. With artificial intelligence you never know. «I do not say that we cannot disconnect that hardware, but I am clear that these things are misleading, the models are really misleading, and as a hardware expert, I want to make sure it from the first moment.»
The evolution of artificial intelligence
Although Richard Ho has not spoken specifically about a general artificial intelligence, a term that is increasingly generated among experts, he has advanced that AI will evolve in the future. And that the agents will work in the background all the time. Something that It will require greater surveillance about them.
«Some might be consulting a tool, others could be looking for a website. Others are thinking, and others need to communicate with each other,» says Ho. To do this, he admits, the challenge will be considerable. And it will be important to be prepared. In fact, if the most catastrophists approach what will happen, it will be vital for all, it seems.
Know How we work in NoticiasVE.
Tags: Artificial intelligence