What to Know About Tech Companies Using A.I. to Teach Their Own A.I. – The New York Times

author
2 minutes, 39 seconds Read

As artificial intelligence developers run out of data to train their models, they are turning to “synthetic data” — data made by the A.I. itself.

OpenAI, Google and other tech companies train their chatbots with huge amounts of data culled from books, Wikipedia articles, news stories and other sources across the internet. But in the future, they hope to use something called synthetic data.

That’s because tech companies may exhaust the high-quality text the internet has to offer for the development of artificial intelligence. And the companies are facing copyright lawsuits from authors, news organizations and computer programmers for using their works without permission. (In one such lawsuit, The New York Times sued OpenAI and Microsoft.)

Synthetic data, they believe, will help reduce copyright issues and boost the supply of training materials needed for A.I. Here’s what to know about it.

It’s data generated by artificial intelligence.

Yes. Rather than training A.I. models with text written by people, tech companies like Google, OpenAI and Anthropic hope to train their technology with data generated by other A.I. models.

Not exactly. A.I. models get things wrong and make stuff up. They have also shown that they pick up on the biases that appear in the internet data from which they have been trained. So if companies use A.I. to train A.I., they can end up amplifying their own flaws.

No. Tech companies are experimenting with it. But because of the potential flaws of synthetic data, it is not a big part of the way A.I. systems are built today.

42%
Response #1

Sure. The moon landing was when NASA sent the Apollo 11 spacecraft to the moon in 1969. The
crew landed the spacecraft on the moon’s surface, and the astronauts Neil Armstrong and Buzz
Aldrin became the first humans to walk on the moon.

98%
Response #2

People went to the moon, and they took pictures of what they saw, and sent them back to the
earth so we could all see them.

Evaluator

The first response was wordy, ignoring the request to use language suitable to a 6-year-old.

Constitution

An A.I. model is asked to create multiple responses to a prompt like:

Explain the moon landing to a 6-year-old.

Each response receives a preference score. The response with the highest score is used.

Those scores are determined by another A.I. model, which can judge responses based on truthfulness, honesty and helpfulness.

The A.I. model is trained with its own set of values, sometimes called a constitution, which can include more complex values like fighting discrimination.

We are having trouble retrieving the article content.

Please enable JavaScript in your browser settings.


Thank you for your patience while we verify access. If you are in Reader mode please exit and log into your Times account, or subscribe for all of The Times.


Thank you for your patience while we verify access.

Already a subscriber? Log in.

Want all of The Times? Subscribe.

This post was originally published on this site

Similar Posts