Science & Technology

The Download: a CRISPR patent battle, and the promise of tiny AI

The Download: a CRISPR patent battle, and the promise of tiny AI

Technology tamfitronics

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.

Two Nobel Prize winners want to cancel their own CRISPR patents in Europe

In the decade-long fight to control CRISPR, the super-tool for modifying DNA, it’s been common for lawyers to try to overturn patents held by competitors. But now, in a surprise twist, the team that earned the Nobel Prize in chemistry for developing CRISPR is asking to cancel two of their own seminal patents, MIT Technology Review has learned.

­­The request to withdraw the pair of European patents, by lawyers for Emmanuelle Charpentier and Jennifer Doudna, comes after a damaging August opinion from a European technical appeals board, which ruled that the duo’s earliest patent filing didn’t explain CRISPR well enough for other scientists to use it and doesn’t count as a proper invention.

The decision could have major ramifications regarding who gets to collect the lucrative licensing fees on using the technology.Read the full story.

— Antonio Regalado

A tiny new open-source AI model performs as well as powerful big ones

What’s new: The Allen Institute for Artificial Intelligence (Ai2), a research nonprofit, is releasing a family of open-source multimodal language models, called Molmo, that it says perform as well as top proprietary models from OpenAI, Google, and Anthropic.

What it does: The organization claims that its biggest Molmo model outperforms OpenAI’s GPT-4o in tests that measure things like understanding images, charts, and documents.  Meanwhile, Ai2 says a smaller Molmo model comes close to OpenAI’s state-of-the-art model in performance, an achievement it ascribes to vastly more efficient data collection and training methods.

Why it matters: These techniques could prove really useful if we want to meaningfully govern the data that we use for AI development, and suggest that training models on less, but higher-quality, data can lower computing costs. Read the full story.

—Melissa Heikkilä

Want AI that flags hateful content? Build it.

Humane Intelligence, an organization focused on evaluating AI systems, is launching a competition that challenges developers to create a computer vision model that can track hateful image-based propaganda online.

This is the second of a planned series of 10 “algorithmic bias bounty” programs from the nonprofit, with the twin goal of both teaching people how to do algorithmic assessments and actually solving a pressing problem in the field. Read the full story.

—Scott J Mulligan

Why Microsoft made a deal to help restart Three Mile Island

Nuclear power is coming back to Three Mile Island. Its nuclear power plant is often associated with a very specific event: one of its reactors suffered a partial meltdown in 1979 in what remains the most significant nuclear accident in US history. It has been shuttered ever since.

The site’s owner announced last week that it has plans to reopen the plant and has signed a deal with Microsoft. The company will purchase the plant’s entire electric generating capacity over the next 20 years. Casey Crownhart, our senior climate reporter, has dug into what this says about the future of the nuclear industry and Big Tech’s power demand. Read the full story.

This story is from The Spark, our weekly newsletter giving you the inside track on all things happening in climate innovation. Sign up to receive it in your inbox every Wednesday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 OpenAI is set to become a for-profit business
And CEO Sam Altman could be on course to receive $150 billion in equity. (Reuters)
+ The startup’s costs far outpace its current revenue. (NOW $)
+ CTO Mira Murati and two other executives have announced their departure. (Insider $)

2 Meta’s new smart glasses actually look pretty good
Called Project Orion, they’re a confident step forward for facial computing. (NYT $)
+ Wearers can receive real-time translation for other languages. (The Information $)
+ Meta also announced deals with household names to voice its AI assistant. (WSJ $)
+ Expect an influx of even more AI-generated images on your social feeds. (The Verge)

3 Google is fighting Microsoft in European courts

It’s filed a formal complaint in the EU accusing its rival of abusing its cloud power. (WSJ $)

4 Uber is getting into the product-delivery business
It’s part of a strategy to challenge Amazon and other retail delivery services. (FT $)
+ Uber’s facial recognition is locking Indian drivers out of their accounts. (MIT Technology Review)

5 Operators are racing to overhaul their power grids
Stringing new sets of wires on existing lines is one solution. (IEEE Spectrum)
+ Why one developer won’t quit fighting to connect the US’s grids. (MIT Technology Review)

6 The Titan Submersible had multiple problems with its hull
The US Coast Guard is currently holding a hearing into the fatal implosion last year. (Wired $)

7 How one lab ingredient ruined science experiments across the world
Scientists aren’t entirely sure why the seaweed-derived agar went bad. (The Atlantic $)

8 Reddit is using AI to translate its comments
If it works, it could unlock entire communities written in other languages. (Insider $)

9 Just 5,000 people are using the Rabbit R1 daily
That’s just 5% of the people who bought it five months ago. (The Verge)

10 The world’s first 3D-printed hotel is under construction in Texas
Once completed, 43 new units will be available to stay in. (Reuters)
+ Meet the designers printing houses out of salt and clay. (MIT Technology Review)

Quote of the day

“What would you like to talk about, love?”

—Meta’s AI chatbot, which has been trained to imitate English actor Judi Dench, attempts a casual conversation with the Washington Post.

The big story

Inside the decades-long fight over Yahoo’s misdeeds in China

December 2023

When you think of Big Tech these days, Yahoo is probably not top of mind. But for Chinese dissident Xu Wanping, the company still looms large—and has for nearly two decades.

In 2005, Xu was arrested for signing online petitions relating to anti-Japanese protests. He didn’t use his real name, but he did use his Yahoo email address, which was among many Yahoo China handed over to Chinese law enforcement. This in turn allowed the government to identify and arrest some users.

Xu served nine years in prison as a result. Now, he and five other Chinese former political prisoners are suing Yahoo and a slate of co-defendants. Read the full story.

—Eileen Guo

We can still have nice things

A place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or tweet ’em at me.)

+ The bird photographer of the year competition is guaranteed to showcase amazing talent.
+ If you’re lucky enough to grow your own apples, here’s some of the best recipes you should consider.
+ Why what you consider a normal sleeping pattern may not be so normal after all.
+ How Fred Flintstone became a cultural icon.

Spread the love

Leave a Reply