AI Is Using Your Likes to Get Inside Your Head

What is The future as the age of artificial intelligence? Max levchin– PayPal sees a new and very valuable role in the COFOUNDER and the CEO, a new and very valuable role in preparing the information to come to the results in accordance with the Ai’s decision.

AI systems are a well-known recognized fluid because the AI ​​systems have caused many different results to improve and reward the performance of a computer provided by a clear award function.

To promote a corrective force, AI developers often use human feedback (RLHF) to learn often. In fact, the computer puts the thumb on a scale, because it reflects the true advantages of real people on the information belonging to a computer model. So how much does he need to enter the preferred information and the input? So far, this has been a problem with RLHF: if human controllers and annotators are required to hire feedback to enter feedback, it is an expensive way.

And this is the problem that Levh’s thinks. This is looking for a sitting stacked source today FacebookAs a god of any developer who wants to train a smart agent for people preferred to people. And how much does this have a contract? “I claim that one of the most valuable items of Facebook owners, the liking of information,” Levchin said. Indeed, in this infection in the development of artificial intelligence, “which content is approved by people, to use it to make AI models, is one of the most valuable things on the Internet.”

While Levchin learned from human choices through the Likes button, AI, these advantages form this advantages first. In fact, social media platforms are not just to predict the likes, to predict the button to show the button itself in an obsolete.

It was a surprising observation, because as most people talked, the forecasts were mainly from another angle, explaining how the EU will affect the performance, but the forecasts that the EU will change the world. We already heard that AI is applied to improve social media algorithms. In the early 2024, for example, experimented with Facebook Using AI Videos to redesign the algorithm that recommends users. Can the variables appear better to predict the user to predict the next one to watch more? This showed that the early test could show the result: to apply the AI ​​to the task paid for a longer period – performance metric Facebook has hoped to increase Facebook.

When we ask YouTube Cofounder Steve said that what the future is a similar button, “Sometimes” Sometimes the button you want to watch according to imaging and distribution rules, it is to make any possible information as easy and accurate as possible. “

However, it continued to manage sharp or temporary changes in terms of needed needs, due to a reason similar key, the need for life occurrences or conditions. “I have days who want to watch the content that are more suitable for a few more, tell me my children,” he said. Chen also involves the role of the similar key to the involvement of the similar button, the other major group of viewers and creators, because of the other major group that acts as the simplest hinge connecting these three groups. With one tap, a follower also proves the evaluation and advantage of the assessment and reviews directly to the content provider and the advantages of the advertiser.