Skip to Content

The Thinking Behind the 10 Breakthrough Technologies

Here’s what we look for, and what we hope to avoid.
February 22, 2017

Every year, MIT Technology Review selects the 10 technologies we believe are the big breakthroughs of the last year: innovations that are a clear advance in their field.

We’ve published a list of the year’s most impactful technologies since 2002. We’re sometimes wrong. We once thought social media and broadcast television would merge (see “Social TV”). But they remain separate streams that people can experience simultaneously, tweeting their impressions of presidential debates as they watch them on TV. More often, we’re not so much wrong as too early: cancer genomics, where doctors sequence the mutations of a patient’s tumor to better match the drugs most likely to help, was less practicable when sequencing was more expensive (see “Cancer Genomics”).

What do we like? We are cross-­disciplinary in our thinking: we enjoy tracing how developments in one field lead to advances in another. A breakthrough in artificial intelligence (see “Deep Learning”) has become crucial to the ambitions for self-driving cars (see “Tesla Autopilot”). We applaud ambitious solutions, such as Google’s plan to bring Internet access to everyone in the world (see “Project Loon”). And we admire elegance and power: we were blown away when scientists used CRISPR to engineer two macaque monkeys, demonstrating the vast potential of gene editing (see “Genome Editing”).

This year’s 10 Breakthrough Technologies reflect the same tastes. “Reinforcement Learning” describes an ambitious approach in AI: computers repeat an action until something difficult goes more smoothly, whereupon the system favors the behavior that led to that outcome. According to senior editor Will Knight, reinforcement learning is an old idea, toyed with by AI pioneers like Marvin Minsky, that never quite worked. But in March 2016, “AlphaGo, a program trained using reinforcement learning, destroyed one of the best Go players of all time … The feat was astonishing, because it is virtually impossible to build a good Go-playing program … Not only is the game extremely complex, but even accomplished players may struggle to say why certain moves are good or bad.” Knight says that reinforcement learning, currently being explored by Uber, OpenAI, and DeepMind, could speed the development of self-driving cars and robots that can reliably grasp objects.

Or consider the cross-disciplinary approach in “Reversing Paralysis”, which combines neuroscience and electronics. Senior editor Antonio Regalado describes how a French neuroscientist named Grégoire Courtine installed a recording device inside the skull of a semi-paralyzed macaque monkey, and then sutured electrodes around the animal’s partially severed spinal cord. “A wireless connection joined the two electronic devices. The result: a system that read the monkey’s intention to move and then transmitted it immediately in the form of bursts of electrical stimulation to its spine.” Suddenly, the monkey’s leg could extend and flex, and it “hobbled forward.” In the past, a few people have controlled robotic arms using brain implants; but by wirelessly connecting brain-reading technologies to electrical stimulators, researchers like Courtine are creating “neural bypasses” that could allow the disabled to walk.

These are just two of the technologies in this year’s list. Read all 10 technologies, and tell me which you think are most remarkable at

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.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

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 with a list of newsletters you’d like to receive.