This Week’s Superior Tech Tales From Across the Internet (By means of April 8)

Builders Are Connecting A number of AI Brokers to Make Extra ‘Autonomous’ AI
Chloe Xiang | Motherboard
“A number of builders are attempting to create an ‘autonomous’ system by stringing collectively a number of situations of OpenAI’s giant language mannequin (LLM) GPT that may do various issues by itself, corresponding to execute a sequence of duties with out intervention, write, debug, and develop its personal code, and critique and repair its personal errors in written outputs.”

AI Is Operating Circles Round Robotics
Jacob Stern | The Atlantic
“AI’s latest progress has been fueled to a major extent by coaching bigger fashions with better computation energy on bigger knowledge units. Roboticists inclined towards this method—hoping to use the identical machine-learning strategies which have proved so fruitful for giant language fashions—run into issues.”

Tiny Hybrid Robotic Can Establish, Seize a Single Cell
Paul McClure | New Atlas
“As soon as the hybrid propulsion system was assembled, researchers have been in a position to exhibit the micro-robot’s capabilities. They used it to seize a single crimson blood cell, most cancers cells, and a single bacterium, demonstrating that the micro-robot may distinguish between a wholesome cell and one which had been broken by a drug or a dying cell and one which was present process a pure ‘suicide’ course of (apoptosis). As soon as captured, the cell might be moved to an exterior instrument for additional evaluation.”

Lab-Grown Burgers Have a Secret Ingredient: Crops
Matt Reynolds | Wired
“Two firms within the US have the Meals and Drug Administration’s nod that their cultivated meat is protected for human consumption, and are awaiting additional sign-off from the Division of Agriculture earlier than they will promote their meat in eating places and shops. However the economics of rising animal cells in bioreactors are nonetheless eye-watering. The best solution to get meat on the market that folks can afford is to mix costly bioreactor-brewed animal cells with less expensive plant-based proteins. The fast way forward for cultivated meat is hybrid.”

This AI Clock Makes use of ChatGPT to Generate Tiny Poems That Inform the Time
James Vincent | The Verge
“ChatGPT has been one of many web’s favourite toys for months now, however individuals are nonetheless discovering novel and enjoyable methods to make use of the AI chatbot. Living proof is that this rhyming E Ink clock created by designer and blogger Matt Webb. It makes use of ChatGPT to create a brief two-line rhyme that additionally tells the time for each minute of the day. It’s unbelievable and we would like one.”

SECURITY

Three Methods AI Chatbots Are a Safety Catastrophe
Melissa Heikkiläarchive web page | MIT Know-how Overview
“Tech firms are racing to embed these fashions into tons of merchandise to assist individuals do every part from e-book journeys to arrange their calendars to take notes in conferences. However the best way these merchandise work—receiving directions from customers after which scouring the web for solutions—creates a ton of recent dangers. With AI, they could possibly be used for all kinds of malicious duties, together with leaking individuals’s personal data and serving to criminals phish, spam, and rip-off individuals. Consultants warn we’re heading towards a safety and privateness ‘catastrophe.’i”

LAW

Secure Diffusion Copyright Lawsuits May Be a Authorized Earthquake for AI
Timothy B. Lee | Ars Technica
“There are some robust arguments that copyright’s honest use doctrine permits Stability AI to make use of the photographs. However there are additionally robust arguments on the opposite aspect. There’s an actual chance that the courts may determine that Stability AI violated copyright regulation on a large scale. That may be a authorized earthquake for this still-nascent business.”

The Name to Halt ‘Harmful’ AI Analysis Ignores a Easy Reality
Sasha Luccioni | Wired
“Tech leaders’ Open Letter proposed a pause on ChatGPT. However researchers already know the right way to make synthetic intelligence safer. …As an alternative of specializing in ways in which AI could fail sooner or later, we should always concentrate on clearly defining what constitutes an AI success within the current.“

Icy Moons With Huge Oceans Are the Newest Candidates for Alien Life
Editorial Employees | The Economist
“All 4 of the photo voltaic system’s fuel giants are both identified or suspected to have watery moons of their very own. There may be even some proof that the identical could also be true for Pluto, a dwarf planet that orbits within the frigid darkness past the orbit of Uranus. Assuming that fuel giants in different star programs even have moons—and there’s no motive to imagine they don’t—that drastically raises the variety of locations within the galaxy through which life may have arisen.”

Picture Credit score: Maxim Berg / Unsplash