This week’s awesome tech stories from around the web (through November 19)

Picture unlimited creativity at your fingertips
Kevin Kelly | Wired
“For the first time in history, people can devise everyday acts of creativity on demand, in real time, at scale, cheaply. Synthetic creativity is now a commodity. The ancient philosophers will turn in their graves, but it turns out that to generate creativity – to create something new – all you need is the right code. We can inject it into tiny devices that are currently inert, or we can apply creativity to large statistical models, or incorporate creativity into drug discovery routines. What else can we use synthetic creativity for?’

MoMA’s newest artist is an AI trained on 180,000 works, from Warhol to Pac-Man
Jesus Diaz | Fast Company
“The colossal installation—a stunning 24-by-24-foot digital screen that fills MoMA’s entire lobby—renders an infinite moving stream of images, each one dreamed up as you watch by an AI model fed by the museum’s entire art collection. That flow is controlled by what’s going on around it, making the piece feel alive.”

New records for the largest and smallest AI computers
Samuel K. Moore | IEEE Spectrum
“Training AI has been a problem that has led to billions of dollars in investment, and it seems to be paying off. “A few years ago we were talking about training these networks in days or weeks, now we’re talking minutes,” says Dave Salvator, director of product marketing at Nvidia.

Meet Midnight, the eVTOL air taxi that Archer will take into production
Los Blaine | New Atlas
“As with all eVTOL companies, Archer promises to be discreet in urban operations, 100 times quieter than a helicopter. It also promises to be much cheaper, at about the price of an Uber-style service per mile, and this ultra-affordable pricing is key to the eVTOL sector’s rapid expansion plans. And the expansion will be rapid indeed. Archer is building an initial plant in Georgia designed to pump out up to 650 aircraft a year. Everything is going according to plan, which will soon be expanded to produce 2,300 per year.”

NASA gets its mojo back with a spectacular nighttime launch of the SLS rocket
Eric Berger | Ars Technica
“The US space agency has not launched a rocket into orbit since 2011, when NASA last flew its historic space shuttle. Furthermore, NASA had not flown a young orbital rocket into space since the shuttle’s debut in 1981. So on Wednesday morning, remarkably, NASA flew its first new rocket in more than four decades.”

Sam Bankman-Fried’s Dangerous Business
Ben Cohen | The Wall Street Journal
“Nobody as rich as Sam Bankman-Fried has ever spent so much time talking to podcasters and explaining how they got rich. Weeks before the collapse of his cryptocurrency exchange and the spectacular collapse of his wealth, the CEO of FTX gave an interview that began with an illuminating question: What was the first thing his company did better than any other? “Manage the risk,” he said.

Intel Unveils Real-Time Deepfake Detector, Claims 96% Accuracy
Sharon Goldman | VentureBeat
“[FakeCatcher] is based on photoplethysmography, or PPG, a method for measuring the amount of light absorbed or reflected by blood vessels in living tissue. When the heart pumps blood, it goes to the veins, which change color. …You can’t see it with your eyes, but it is computationally visible. [Intel’s Ilke] Demir said VentureBeat. PPG signals are known, but have not been applied to the deepfake problem before.I

Will Physics Prevent SpinLaunch’s Success?
Ethan Siegel | Big Think
“Although humanity has managed to send spacecraft into orbit and even beyond Earth’s gravitational pull, the only way we’ve done it is through fuel-guzzling rocket launches. In the past, alternatives have been proposed: rail guns, missile launchers, space elevators and more, but none have ever delivered a single payload into orbit. With a working prototype successfully launching objects at 1,000 miles per hour, SpinLaunch looks promising. But will the laws of physics stand in the way of a full-scale version?”


Binance has a cunning plan to save crypto—but it’s too late
Joel Khalili and Morgan Meaker | Wired
“None of the measures implemented by crypto exchanges will stave off the period of increased regulatory scrutiny that is now expected to begin. To date, efforts to regulate crypto companies have moved very slowly, partly as a result of the complexity of the underlying technology, says Charley Cooper, former COO of the Commodity Futures Trading Commission (CFTC) in the US. But the scale of the FTX collapse is likely to light a fire under regulators around the world.”

Because Meta’s latest model with a big tongue only survived three days online
Will Douglas Heaven | MIT Technology
“Meta’s mistake—and its hubris—shows once again that Big Tech has a blind spot about the serious limitations of big language models. There is a large body of research that highlights the flaws of this technology, including its tendencies to reproduce biases and endorse lies as fact. However, Meta and other companies working on large language models, including Google, did not take it seriously.”

Image credit: lilzidesigns / Unsplash

Leave a Reply

Your email address will not be published. Required fields are marked *