Skip to Content

Is Artificial Intelligence Stuck in A Rut?

The former director of Uber’s AI lab says the field is in danger of losing sight of its long-term goals.
March 29, 2017

Don’t be fooled by all the money pouring into artificial intelligence projects these days. The former head of Uber’s AI Lab, Gary Marcus, warns that the field isn’t moving nearly as fast as many people think.

Speaking at MIT Technology Review’s EmTech Digital conference in San Francisco this week, Marcus said the current obsession with statistical machine learning, combined with the short-term focus of companies that are investing in AI, was limiting progress towards human-level artificial intelligence.

“My biggest fear is not Skynet,” Marcus said, referring to the killer artificial intelligence in the Terminator films. “It’s getting stuck.”

Marcus argued that despite recent technical advances, there are many simple things that computers cannot do, and that these limitations are holding back efforts to make progress towards real general intelligence.

Marcus is a well-known critic of AI hype, but his comments contrasted starkly with those of Ilya Sutskever, the research director of OpenAI, a non-project created by Elon Musk and others to do fundamental and open research on AI.

In an earlier talk, in which Sutskever described a new approach to machine learning, he suggested that general, or human-level, AI might not be so far away.

“[It] seems far off right now but [was] way more far off five years ago,” he said. “The number of people and the amount of effort going into developing these algorithms is extremely high—things are moving forward at a very healthy pace.”

Marcus, who left Uber earlier this month to spend more time with his family, and who is also a professor at NYU, said during his talk that corporate investment in AI might not be such a good thing for the field’s long-term goals. In times past, a lack of progress caused investment in AI to dry up. This time around, Marcus suggested, too much investment might cause researchers to lose sight of the long-term goals.

Asked if the commercial interest in AI, which is pulling many academics into industry, was a good thing, Marcus, said it was a double-edged sword. “Certain things can be done at a company like Uber, they have enormous resources,” Marcus said. “The problem with the corporate market is there’s a short-term focus. How do you make money out of deep learning today?”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.