In brief OpenAI has disbanded its AI robotics team and is no longer trying to apply machine learning to physical machines.
Wojciech Zaremba, co-founder of OpenAI, who led the robotics group confirmed that the company recently broke up the team to focus working on more promising areas of artificial general intelligence research.
“Here’s a reveal … as of recently we changed the focus at OpenAI, and I actually disbanded the robotics team,” he said during an episode of the Weights & Biases podcast.
Zaremba said a lack of training data was holding the robotics research back: there wasn’t enough information on hand to teach the systems to the level of intelligence desired.
“From the perspective of what we want to achieve, which is to build AGI, I think there was actually some components missing,” he added. A spokesperson from OpenAI this week confirmed it had, indeed, stopped working on robotics.
You can now download a top AI model for protein prediction
DeepMind released AlphaFold, the most advanced protein-structure-predicting machine-learning model yet, on GitHub, this week.
If you want to play around with it, you’ll need to be familiar with Docker, and have the space to store hundreds of gigabytes of genetic sequencing data as well as the model.
AlphaFold is trained to predict how a protein folds and takes shape given its constituent amino acids. Last year, DeepMind entered its system into the Critical Assessment of Protein Structure Prediction contest, and thrashed its rivals.
DeepMind’s goal is to get the model accurate enough to be useful in developing drugs that can target specific proteins to cure or mitigate diseases. A paper by DeepMind on AlphaFold’s design was published this month in Nature.
In a separate project, a large team of researchers at various universities and academic institutions also published their own open-source AI protein folding model. Known as RoseTTaFold, it doesn’t perform as well as AlphaFold though it’s not too shabby, according to a paper published in Science.
Is it wrong to bring dead people back to life in documentaries without telling the audience?
A New Yorker review of Roadrunner, a documentary about the late and great Anthony Bourdain, has sparked questions over whether it’s ethical or not to use machine-learning technology to low-key fake people’s voices.
In the magazine piece, the documentary’s filmmaker Morgan Neville admitted to using software that mimicked Bourdain’s voice, making the celebrity chef and writer say words he had only written. Specifically, the software was used to read out an email Bourdain had written to a friend. The code was trained on clips of Bourdain speaking on TV, radio, audiobooks, and podcasts.
“If you watch the film … you probably don’t know what the other lines are that were spoken by the AI, and you’re not going to know,” Neville said. “We can have a documentary-ethics panel about it later.”
Should the director tell viewers or listeners when an audio clip has been synthetically generated? Does it matter, seeing as Bourdain did express those sentiments albeit in an email and not into a microphone? Will this blow a hole in trust in future documentaries, journalism, and media output? This Tech Policy Press interview with Sam Gregory – a deep-fakes expert and program director of Witness Media Lab – has more on that.
Discord believes AI can help moderate hate speech online
IRC-for-the-next-generation Discord has snapped up an AI startup for its automated moderation tools.
Sentropy, based in Palo Alto, California, confirmed the deal in a blog post this week. “Three years after starting this company with Michele, Ethan, and Taylor, I’m thrilled to announce that we’re joining Discord to continue fighting against hate and abuse on the internet,” said CEO and co-founder John Redgrave.
The upstart has built proprietary machine-learning models said to be capable of detecting hate speech and toxic language to shut down online harassment. The amount Discord paid to acquire Sentropy’s technology and team was not disclosed.
Discord was known for being primarily popular with gamers, though it has exploded in use in other communities, from programming to cryptocurrencies. It reportedly walked away from a $10bn offer by Microsoft, earlier this year. ®
The US Ninth Circuit Court of Appeals on Friday upheld a lower court’s refusal to block California’s net neutrality law (SB 822), affirming that state laws can regulate internet connectivity where federal law has gone silent.
The decision is a blow to the large internet service providers that challenged California’s regulations, which prohibit network practices that discriminate against lawful applications and online activities. SB 822, for example, forbids “zero-rating” programs that exempt favored services from customer data allotments, paid prioritization, and blocking or degrading service.
In 2017, under the leadership of then-chairman Ajit Pai, the US Federal Communications Commission tossed out America’s net neutrality rules, to the delight of the internet service providers that had to comply. Then in 2018, the FCC issued an order that redefined broadband internet services, treating them as “information services” under Title I of the Communications Act instead of more regulated “telecommunications services” under Title II of the Communications Act.
California lawmaker Scott Wiener (D) crafted SB 822 to implement the nixed 2015 Open Internet Order on a state level, in an effort to fill the vacuum left by the FCC’s abdication. SB 822, the “California Internet Consumer Protection and Net Neutrality Act of 2018,” was signed into law in September 2018 and promptly challenged.
In October 2018, a group of cable and telecom trade associations sued California to prevent SB 822 from being enforced. In February, 2021, Judge John Mendez of the United States District Court for Eastern California declined to grant the plaintiffs’ request for an injunction to block the law.
So the trade groups took their case to the Ninth Circuit Court of Appeals, which has now rejected their arguments. While federal laws can preempt state laws, the FCC’s decision to reclassify broadband services has moved those services outside its authority and opened a gap that state regulators are now free to fill.
“We conclude the district court correctly denied the preliminary injunction,” the appellate ruling [PDF] says. “This is because only the invocation of federal regulatory authority can preempt state regulatory authority.
The FCC no longer has the authority to regulate in the same manner that it had when these services were classified as telecommunications services
“As the D.C. Circuit held in Mozilla, by classifying broadband internet services as information services, the FCC no longer has the authority to regulate in the same manner that it had when these services were classified as telecommunications services. The agency, therefore, cannot preempt state action, like SB 822, that protects net neutrality.”
The Electronic Frontier Foundation, which supported California in an amicus brief, celebrated the decision in a statement emailed to The Register.
“EFF is pleased that the Ninth Circuit has refused to bar enforcement of California’s pioneering net neutrality rules, recognizing a very simple principle: the federal government can’t simultaneously refuse to protect net neutrality and prevent anyone else from filling the gap,” a spokesperson said.
“Californians can breathe a sigh of relief that their state will be able to do its part to ensure fair access to the internet for all, at a time when we most need it.”
There’s still the possibility that the plaintiffs – ACA Connects, CTIA, NCTA and USTelecom – could appeal to the US Supreme Court.
In an emailed statement, the organizations told us, “We’re disappointed and will review our options. Once again, a piecemeal approach to this issue is untenable and Congress should codify national rules for an open Internet once and for all.” ®
An existing drug called PARP inhibitor can be used to exploit a vulnerability in the way breast cancer cells repair their DNA, preventing spread to the brain.
For a long time, there have been limited treatment options for patients with breast cancer that has spread to the brain, sometimes leaving them with just months to live. But scientists at the Royal College of Surgeons Ireland (RCSI) have found a potential treatment using existing drugs.
By tracking the development of tumours from diagnosis to their spread to the brain, a team of researchers at RCSI University of Medicine and Health Sciences and the Beaumont RCSI Cancer Centre found a previously unknown vulnerability in the way the tumours repair their DNA.
An existing kind of drug known as a PARP inhibitor, often used to treat heritable cancers, can prevent cancer cells from repairing their DNA because of this vulnerability, culminating in the cells dying and the patient being rid of the cancer.
Prof Leonie Young, principal investigator of the RCSI study, said that breast cancer research focused on expanding treatment options for patients whose disease has spread to the brain is urgently needed to save the lives of those living with the disease.
“Our study represents an important development in getting one step closer to a potential treatment for patients with this devastating complication of breast cancer,” she said of the study, which was published in the journal Nature Communications.
Deaths caused by breast cancer are often a result of treatment relapses which lead to tumours spreading to other parts of the body, a condition known as secondary or metastatic breast cancer. This kind of cancer is particularly aggressive and lethal when it spreads to the brain.
The study was funded by Breast Cancer Ireland with support from Breast Cancer Now and Science Foundation Ireland.
It was carried out as an international collaboration with the Mayo Clinic and the University of Pittsburgh in the US. Apart from Prof Young, the other RCSI researchers were Dr Nicola Cosgrove, Dr Damir Varešlija and Prof Arnold Hill.
“By uncovering these new vulnerabilities in DNA pathways in brain metastasis, our research opens up the possibility of novel treatment strategies for patients who previously had limited targeted therapy options”, said Dr Varešlija.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
Microsoft’s second attempt at its interesting dual-screen Android smartphone corrects some mistakes of the original, but falls short of a revolution due to a series of oddities created by its physical laptop-like form.
Looking more like a tiny convertible computer than a phone, the Surface Duo 2 starts at £1,349 ($1,499/A$2,319), a lot for a regular smartphone but slightly cheaper than folding-screen rivals.
It opens like a book, with each half just 5.5mm thick, and a hinge that allows it to fold all the way over.
Inside are a pair of 90Hz OLED screens each measuring 5.8in on the diagonal. They can be used on their own or combined as one display measuring 8.3in – a similar size to an iPad mini. Both screens are covered in traditional scratch-resistant smartphone glass and have large, old-fashioned bezels top and bottom.
Having two separate displays rather than one that folds in half creates a major drawback: a gap in the middle of the screen big enough that you can see through it, which is much harder to ignore than the crease in the middle of a flexible display as found on the Samsung Galaxy Z Fold 3.
You can use two different apps at the same time on the two screens. The theory is sound, but I found few pairings were useful beyond simple messaging apps and a browser. More useful was using one screen for a note-taking app and the other for a full keyboard like a mini laptop.
Some apps spanned across both displays, like Outlook, can put different information on each screen, such as your inbox on one side and an open message on the other. Some games, including Asphalt 9 and Microsoft’s Xbox Game Pass streaming service, put controls on one screen and the action on the other. But there are very few apps and games optimised for this setup.
Connectivity: 5G, USB-C, wifi 6, NFC, Bluetooth 5.1 and location
Water resistance: IPX1 (dripping water)
Dimensions closed: 145.2 x 92.1 x 11.0mm
Dimensions open: 145.2 x 184.5 x 5.5mm
2021’s top Android chip
The Duo 2 has last year’s top Qualcomm Snapdragon 888 chip with 8GB of RAM, matching the performance of top-flight Android smartphones from 2021 and capable of running two apps running side-by-side without slowdown.
Battery life is more variable than a traditional phone. It lasts about 32 hours between charges, with both screens used for about four hours with a variety of messaging, browsing and work apps. It lasts about a third longer if you mostly use only one screen. That’s a considerably shorter battery life than a regular smartphone and behind the Z Fold 3.
Microsoft does not provide an expected lifespan for the Duo 2’s battery; those in similar devices typically maintain at least 80% of their original capacity for in excess of 500 full charge cycles. Microsoft charges an out-of-warranty service fee of £593.94 to repair devices and £568.44 to replace the battery. The previous generation Surface Duo scored only two out of 10 on iFixit’s repairability scale.
The Duo 2 runs Android 11 – not the latest Android 12 – and generally behaves like a standard Android smartphone or tablet with a few small additions that make it easier to use each screen separately. One of the best is the ability to drag the gesture bar at the bottom of an app to move it between screens or to drop it on to the gap between the screens to span it across both displays.
The software can be a bit unpredictable at times, such as opening the keyboard or text box of an app on another screen or hiding a second app from the screen when you try to type. But it is generally a fast and responsive experience given how unusual the device is.
The Duo 2 will receive three years of software updates from release, including monthly security patches, which is disappointingly at least a year short of what rivals, including Samsung and Apple, offer. Microsoft’s last planned update for the Duo 2 will be 21 October 2024.
The Duo 2 has a triple camera on the back and a 12-megapixel selfie camera above the right-hand screen.
The rear main 12MP camera and 2x telephoto cameras are good, capable of producing detailed shots in a range of lighting conditions. The 16MP ultra-wide camera is reasonable, but a bit soft on detail and struggles with challenging scenes. The camera app has most of the features you’d expect, such as portrait mode, night mode and slow-mo video, and can shoot regular video at up to 4K at 60 frames a second.
The 12MP selfie camera is capable of shooting detailed photos even in middling light, and has access to the dedicated night mode when it gets dark.
Overall, the camera system on the Duo 2 is solid, but it can’t hold a candle to the best in the business.
The Duo 2 supports Microsoft’s Slim Pen stylus, which can be magnetically stored and charged on the back of the device when not in use.
The stereo speakers are decently loud but a bit tinny, fine for watching YouTube videos.
The width of the device makes it a challenge to fit into smaller pockets.
The Surface Duo 2 costs £1,349 ($1,499/A$2,319) with 128GB, £1,429 ($1,599/A$2,469) with 256GB or £1,589 ($1,799/A$2,769) with 512GB of storage.
The Surface Duo 2 is an improvement on its predecessor, but is still a very odd proposition that’s neither a good phone nor a good tablet.
The individual screens are short and stout, forcing lots of scrolling in apps when using it like a phone and making one-handed use very difficult. The gap at the hinge makes combining them into one big tablet screen awkward too.
Using two apps side-by-side works well, but few combinations proved useful or faster than just quick switching between two apps on one screen on a normal phone. There is more potential in apps like Outlook that provide a multi-pane view, but few apps or games are optimised for the dual-screen system.
Microsoft is only offering a disappointing three years of software and security updates from release for the Duo 2, too, losing it a star.
It is good to see Microsoft trying something different. But ultimately the Duo 2’s two screens are just not yet as good or useful as either a single phone screen or a bigger folding screen, making it an expensive halfway house.
Pros: two screens, two apps side-by-side, multiple modes, top performance, hardened glass screens, decent camera, head-turning design.
Cons: gap between screens, few optimised apps, average battery life, bulky camera lump, chunky in pocket, hard to use one-handed, no real water resistance, only three years of software updates from release.