One of the leading experts in artificial intelligence (AI), Gary Marcus, professor emeritus of psychology and neural sciences at New York University, think it will be necessary apply watermarks to bot-generated imagesand something similar in the texts, to distinguish them from human-generated texts and to prevent the spread of misinformation.

In a New York Times article, Marcus, author of two books devoted to AI, tries to avoid the catastrophic tone “I’m not one of those who think the entire planet will be dominated by robots,” but admits that he worries about “what negative actors can do with these things, because no one controls them.”

“Do the benefits outweigh the costs? -he wonders- The most honest answer is that we don’t know. Some of us would like this slow down because we see your costs daily, but that doesn’t mean there aren’t benefits”, he qualifies, especially on the technical side, and incidentally warns against those who advocate ChatGPT as a means of winning some kind of race with China.

With these clues you can recognize when an image was created by artificial intelligence

As an immediate remedy, Marcus suggests some kind of international governance over the world of AI: “What we need it is something global, neutral, non-profit, in which governments and companies participate. We need to coordinate our efforts to create rules (…) We need to create a code of what is acceptable and who is responsible”.

This supranational body would respond to well-founded doubts that have already arisen, giving the following examples: What if the chatbots lie a lot?Is it allowed, who is responsible? as misinformation spreads widely, To do? who bears the costs? Do we want companies to put money into creating tools that detect the disinformation they themselves create?

The expert also warns against the absolute criterion of effectiveness when accepting the massive use of AI as an element to reduce costs: “These tools can give us huge benefits in terms of productivity, but they can also destroy the social fabric. (referring to any written-off jobs). If so, they are no longer worth it,” he proclaims.

Spain supports the development of artificial intelligence, but in accordance with democracy

If rapid regulation of AI and chatbots isn’t cutting edge, Marcus gives an example of what that could mean in the next US presidential election:If we do nothing, no one will believe anything, and all those things they don’t want to believe, they will reject and attribute to the AI. So the problems of polarization and public discourse will only get worse”.

Marcus’s warnings add to those of other experts over the past few days warn of the spectacular rise of AI, virtually uncontrolled until now.