What’s the news? Deep learning is behind the most high-profile successes of machine learning. But this incredible achievement comes with a price: training deep learning models takes a lot of energy. New research now shows how scientists who use cloud platforms to train algorithms can dramatically reduce the energy they use, and thus the emissions they create.
How can they do it?: Simple changes to cloud settings are key. Researchers have created a tool that measures the electricity consumption of any machine learning program running on Azure, Microsoft’s cloud service, at every stage of their project. They estimated the emissions based on the zip codes of servers with 11 machine learning models, and found that they could be significantly reduced if researchers adjusted the settings to use servers in specific geographic locations and at certain times of the day.
The bigger picture: Getting people to sign up to adjust their own settings is an uphill battle. Only 13% of Azure users running machine learning programs have viewed the energy metering tool since its launch in October, so the next step will convince the rest. Read the full story.
The world will need dozens of groundbreaking climate technologies over the next decade
We live in a crucial decade. By 2030, global emissions should be halved, mainly through the massive deployment of existing technologies such as wind turbines, solar panels and electric vehicles. But emerging climate technologies should also come on the market in this decade, even if they will not immediately make a big difference in emissions.
Those on this year’s list of MIT Technology Review Innovators under 35 are seizing the opportunity to decarbonize the economy and make the clean energy transition affordable. Learn about their work and what it takes to help them succeed in this essay by Varun Sivaram, the senior director for clean energy and innovation for US Presidential Special Envoy for Climate, John Kerry.
This essay is part of MIT Technology Review’s 2022 Innovators Under 35 package, which recognizes the most promising young people working in technology today. View the full list here.
The must reads
I’ve scoured the internet to find today’s funniest/important/scary/fascinating stories about technology.
1 The Leaked Data Of A Billion Chinese People Was Online For More Than A Year
It sat undetected in an unsecured database before a hacker offered to sell it. †CNN†
2 The Large Hadron Collider Helped Scientists Find Three New Particles
The combinations have never been seen before. †Motherboard†
+ Don’t fall for these misconceptions about the Hadron Collider’s capabilities. †Big Thinking†
3 How Wall Street emerged unscathed from the crypto carnage
It turns out that regulation is quite useful. †NYT †
+ And it comes for crypto too. †wired †
+ The crypto crash could be a setback for web3. †FT †
Venture capitalists burned by the past decade of insane growth are cautious. †Motherboard†
+ Black investors suffer the most. †FT †
+ It’s okay to opt out of the crypto revolution. (MIT Technology Review)
4 Europe has green-lit Big Tech regulations
However, it will be some time before new laws come into effect. †axios†
+ The UK’s online safety law has been amended to prioritize the detection of child abuse material. †the guard†
The EPA Supreme Court ruling last week doesn’t bode well for US regulation. †protocol†
5 Microsoft is still using emotion-detecting AI
For an app for people with vision loss, despite widespread skepticism about the accuracy of the technology. †protocol†
Emotion AI researchers say exaggerated claims give their work a bad name. (MIT Technology Review)
6 How technology is saving Sri Lanka’s ravaged tourism industry
Includes virtual leopard safaris. †Rest of the world†
7 People shouldn’t hibernate
But a handful of cases suggest that it is possible to enter a state of stupor. †CNET†
8 Everything is a vibe these days
Suggesting it could be time for a vibe shift – away from the vibes themselves. †The Atlantic Ocean †
9 Sports in space is coming
No gravity? No problem. †WSJ †
+ Can constant acceleration be used to produce artificial gravity in space? (MIT Technology Review)
10 The Existential Grief Of Robots
Maybe it’s time to stop projecting our own emotions onto them. †the guard†
+ How we think about robots that feel. (MIT Technology Review)
That said, they are pretty good surgeons. †Spectrum IEEE†