This week’s awesome tech stories from around the web (through July 13)

OpenAI Reportedly Nearing Breakthrough with AI ‘Reasoning’, Reveals Framework for Progress
Benj Edwards | Ars Technica
“(Under the new AGI OpenAI framework) a level 2 artificial intelligence system would supposedly be capable of basic problem solving on par with a human with a PhD but no access to external tools. During the all-hands meeting, OpenAI management reportedly showcased a research project using their GPT-4 model, which scientists believe shows signs of approaching this human-like reasoning ability, according to someone familiar with the discussion who spoke to Bloomberg.”

How AI revolutionized protein science, but it’s not over
Yasemin Saplakoglu | Quanta
“Three years ago, Google’s AlphaFold delivered the biggest AI breakthrough in science to date, accelerating molecular research and prompting deep questions about why we do science. …’The field of protein biology is ‘more exciting right now than it was before AlphaFold,’ Perrakis said. The excitement comes from the promise of revitalizing structure-based drug discovery, accelerating hypothesis generation and the hope of understanding the complex interactions taking place in cells.”

New Fiber Optics technology breaks record for data transfer speed
Margo Anderson | IEEE spectrum
“An international team of researchers has broken the world record for fiber optic communication via commercial fiber. By extending the fiber optic communications bandwidth, the team achieved data transfer speeds four times faster than existing commercial systems – and 33 percent better than the previous world record.

‘Superhuman’ Go AIs still have trouble defending against these simple exploits
Kyle Orlando | Ars Technica
“In an ancient Chinese game Gostate-of-the-art AIs have generally been able to beat the best human players since at least 2016. But in the past few years, researchers have discovered flaws in these top AIs. Go algorithms that give people a fighting chance. By using unorthodox ‘cycle’ strategies – ones that even a novice player could detect and defeat – a cunning human can often exploit loopholes in a top AI’s strategy and trick the algorithm into losing.”

Google creates self-replicating life from digital ‘primordial soup’
Matthew Sparkes | The new scientist
“A self-reproducing artificial life form has emerged from a digital ‘primordial soup’ of random data, despite the lack of explicit rules or goals to support such behavior. The researchers believe that it is possible that more sophisticated versions of the experiment could yield more advanced digital organisms, and if they do, the findings could shed light on the mechanisms behind the origin of biological life on Earth.”

How good is ChatGPT at coding, really?
Michelle Hampson | IEEE spectrum
“Programmers have spent decades writing code for AI models, and now, in full swing, AI is being used to write code. But how does an AI code generator compare to a human programmer? (The new study shows that ChatGPT has an extremely wide range of success when it comes to generating working code — with success rates ranging from 0.66 percent to 89 percent — depending on the difficulty of the task, the programming language, and a number of other factors.”

OpenAI expects AI model costs to drop as adoption surges
Shubham Sharma | VentureBeat
“We introduced GPT-4, the first version, about 15 months ago. Since then, the cost per token/word on the model has decreased by 85-90%. There is no reason why this trend should not continue,” said Olivier Godement, (OpenAI) API Product Lead. …He expects the company’s work on affordability, including efforts to optimize costs at both the hardware and derivation levels, to continue, leading to further declines in the cost of running frontier AI models – similar to what happened with smartphones and TVs. “

Watch these supernovae in (time-lapse) motion
Dennis Overby | The New York Times
“This spring, astronomers who operate Chandra combined its X-ray images into videos that document the evolution of two astrophysical landmarks: the Crab Nebula in the constellation Taurus and Cassiopeia A, a gas bubble and radio noise center in the constellation. Cassiopeia. The videos show twisting, drifting ribbons of the star’s remnants, being tossed about by shock waves and irradiated by radiation from the dense, rotating cores left behind.”

Image credit: Bolivia Inteligente / Unsplash

Leave a Comment

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

Scroll to Top