Experts remain split on whether AI will help or harm humanity

In short: Experts are still sharply divided on whether AI will mostly benefit society or create extreme risks, and there is no clear agreement.

What's going on

Some tech leaders say AI will be a major positive force. Investor Marc Andreessen has argued that AI could be one of the best things humans have made and that it will not end civilization. Supporters point to real progress, like AI systems that helped predict how proteins fold, which can speed up medicine research and helped earn a 2024 Nobel Prize in Chemistry.

These optimists also point to popular tools such as ChatGPT, which can write and answer questions in ways that often feel human. They say AI could help with everyday problems, from tutoring to drafting documents, and even offer companionship for people who feel lonely or depressed.

Others warn about the long-term danger if AI becomes AGI, short for artificial general intelligence (an AI that can learn and think across many tasks like a person). Some “AI doomers” argue that an AGI could end up working against human interests. Geoffrey Hinton, an AI pioneer, has said today’s models may already grasp meaning, and he expects them to gain stronger reasoning skills.

A middle group says both sides are overconfident. Researcher Yann LeCun argues that current AI mostly predicts the next word based on patterns in data, not true understanding. They also note that AI still makes obvious mistakes, like a widely discussed case where Google’s system suggested people eat rocks.

What to watch

The biggest unknown is how fast AI improves and how people choose to use it. Like electricity, the same tool can power hospitals or cause harm if misused. Rules, testing, and day to day choices by companies and governments will likely shape what happens next.

Source: NYTimes