This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.
This new data poisoning tool lets artists fight back against generative AI
What’s happening: A new tool lets artists make invisible changes to the pixels in their art before they upload it online so that if it’s scraped into an AI training set, it can cause the resulting model to break in chaotic and unpredictable ways.
Why it matters: The tool, called Nightshade, is intended as a way to fight back against AI companies that use artists’ work to train their models without the creator’s permission. Using it to “poison” this training data could damage future iterations of image-generating AI models, such as DALL-E, Midjourney, and Stable Diffusion, by rendering some of their outputs useless.
How it works: Nightshade exploits a security vulnerability in generative AI models, one arising from the fact that they are trained on vast amounts of data—in this case, images that have been hoovered from the internet. Poisoned data samples can manipulate models into learning, for example, that images of hats are cakes, and images of handbags are toasters. And it’s almost impossible to defend against this sort of attack currently. Read the full story.
—Melissa Heikkilä
If you’re interested in reading more about poisoning AI datasets, Melissa digs deeper into the story in this week’s edition of The Algorithm, our weekly AI newsletter. Sign up to receive it in your inbox every Monday.
Heat-storing batteries are scaling up to solve one of climate’s dirtiest problems
The news: Antora Energy, a thermal-battery startup, has unveiled its plan to build its first large-scale manufacturing facility in San Jose. The announcement is a big step forward for heat batteries, an industry seeking to become a major player in the energy storage sector.
Why it’s a big deal: Antora’s batteries store renewable energy as heat, which can then be used to manufacture industrial products like cement or glass. Producing industrial heat accounts for about 20% of all global energy demand, in part because this process relies heavily on fossil fuels to generate the high temperatures needed to make products. Antora hopes its heat batteries will help wean heavy industries off fossil fuels and reduce their carbon footprint. Read the full story.
—June Kim
Death to captchas
Earlier this year, HBO Max users hoping to sign in to the service had to pass an audio challenge in which they listened to a bunch of tunes and had to select the one with a repeating pattern.
Websites use these kinds of captchas (the name comes from “Completely Automated Public Turing test to tell Computers and Humans Apart”) to tell whether a user is human or machine. You’ve likely noticed they have only gotten more difficult and more involved.
That’s because of what happens after we solve a captcha: the data from our efforts to label those blurry grids of traffic lights, text, or buses is used to train AI systems, which then get better at defeating captchas, tricking systems into thinking they are human. But both tech firms and consumers alike feel it’s time for a change. Read the full story.
—Shubham Agarwal
This piece is from the next magazine edition of MIT Technology Review, set to go live tomorrow. It’s all about society’s hardest problems, and how we should tackle them. If you don’t already, subscribe now to get a copy when it lands!
2023 Climate Tech Companies to Watch: NotCo and its plant-based foods
Chilean company NotCo is using AI to design and produce plant-based alternatives to dairy and meat products, including vegan burgers, mayonnaise, and chicken that taste more like the real thing.
It has teamed up with some of the world’s biggest brands to help them create foods that have less of an impact on the climate. Read more about NotCo, and check out the rest of the list of Climate Tech Companies to Watch.
The must-reads
I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.
1 A face search engine has blocked searches for children’s faces
But it’s not clear how enforceable the new policy will be. (NYT $)
+ The movement to limit face recognition tech might finally get a win. (MIT Technology Review)
2 The EU may fail to pass its AI Act this year
It’s holding a meeting today to try and speed things up, but there are no guarantees. (Reuters)
+ Five big takeaways from Europe’s AI Act. (MIT Technology Review)
3 Apple’s “carbon neutral” claims are under investigation
The EU isn’t convinced that using credits to cancel out emissions amounts to much more than greenwashing. (FT $)
+ The climate solution actually adding millions of tons of CO2 into the atmosphere. (MIT Technology Review)
4 This week marks a year of Elon Musk’s ownership of X
And very few of us would describe it as a resounding success. (Slate $)
+ Former workers aren’t exactly impressed, either. (Fast Company $)
5 How AI is slowly but surely saving San Francisco
Like it or not, the city is still home to a lot of the buzziest startups. (WP $)
+ College students are ditching their studies for AI’s siren call. (WSJ $)
6 The era of platforms is over
Cross-posting is where it’s at, now. (The Verge)
+ How to fix the internet. (MIT Technology Review)
7 The moon is 40 million years older than we thought 🌒
According to a new analysis of crystals collected during a 1972 moon mission. (CNN)
+ Fossilized molecules are also revealing ancient Earth’s secrets. (Quanta Magazine)
+ What’s next for the moon. (MIT Technology Review)
8 Why fungi plays such an important part in climate change
We’re learning more about how carbon we previously thought of as stored in trees, may be in fungi instead. (The Atlantic $)
9 Inside the rise and rapid fall of Cameo
It turns out novel celebrity videos are exactly that—a novelty. (NYT $)
10 Parental matchmaking is coming to Tinder
But don’t worry, they can’t message any potential love interests. (The Guardian)
Quote of the day
“There are more regulations on sandwich shops than there are on AI companies.”
—Stuart Russell, professor of computer science at the University of California, Berkeley, puts the lack of AI rules into stark relief, the Guardian reports.
The big story
Generative AI is changing everything. But what’s left when the hype is gone?
December 2022
It was clear that OpenAI was on to something. In late 2021, a small team of researchers was playing around with a new version of OpenAI’s text-to-image model, DALL-E, an AI that converts short written descriptions into pictures: a fox painted by Van Gogh, perhaps, or a corgi made of pizza. Now they just had to figure out what to do with it.
Nobody could have predicted just how big a splash this product was going to make. The rapid release of other generative models has inspired hundreds of newspaper headlines and magazine covers, filled social media with memes, kicked a hype machine into overdrive—and set off an intense backlash from creators.
The exciting truth is, we don’t really know what’s coming next. While creative industries will feel the impact first, this tech will give creative superpowers to everybody. The problem is, these models still have no idea what they’re doing. Read the full story.
—Will Douglas Heaven
We can still have nice things
A place for comfort, fun and distraction in these weird times. (Got any ideas? Drop me a line or tweet ’em at me.)
+ The Prince of Darkness himself, Ozzy Osbourne, would love this inventive Crazy Train cover.
+ Margaret Atwood takes a look at an AI-generated Margaret Atwood story—and she is not impressed.
+ The Brits are at it again—this time, it’s playing conkers (thanks Charlotte!)
+ Dissolving paint, bright colors, one utterly hypnotic video.
+ Tell me more about this delicious-looking Bombay sandwich? 🥪