This Week’s Superior Tech Tales From Across the Internet (Via September 27)


Tech

OpenAI and Nvidia’s $100B AI Plan Will Require Energy Equal to 10 Nuclear ReactorsBenj Edwards | Ars Technica

“Nvidia CEO Jensen Huang advised CNBC that the deliberate 10 gigawatts equals the ability consumption of between 4 million and 5 million graphics processing models, which matches the corporate’s whole GPU shipments for this 12 months and doubles final 12 months’s quantity.”

ARTIFICIAL INTELLIGENCE

Spending on AI Is at Epic Ranges. Will It Ever Pay Off?Eliot Brown and Robbie Whelan | The Wall Avenue Journal

“This week, consultants at Bain & Co. estimated the wave of AI infrastructure spending would require $2 trillion in annual AI income by 2030. By comparability, that’s greater than the mixed 2024 income of Amazon, Apple, Alphabet, Microsoft, Meta, and Nvidia, and greater than 5 occasions the dimensions of your complete international subscription software program market.”

Robotics

There Are Extra Robots Working in China Than the Remainder of the World MixedMeaghan Tobin and Keith Bradsher | The New York Occasions

“There have been greater than two million robots working in Chinese language factories final 12 months, based on a report launched Thursday by the Worldwide Federation of Robotics, a nonprofit commerce group for makers of commercial robots. Factories in China put in practically 300,000 new robots final 12 months, greater than the remainder of the world mixed, the report discovered.”

Biotechnology

Huntington’s Illness Breakthrough: What to Know In regards to the Gene RemedyGrace Wade | New Scientist

“An experimental gene remedy has change into the primary therapy to efficiently sluggish the development of Huntington’s illness. Whereas the findings are nonetheless preliminary, the method may very well be a significant breakthrough and should even result in new therapies for different neurodegenerative situations, like Parkinson’s and Alzheimer’s.”

ROBOTICS

Google DeepMind Unveils Its First ‘Considering’ Robotics AIRyan Whitwam | Ars Technica

“Generative AI techniques that create textual content, photos, audio, and even video have gotten commonplace. In the identical approach AI fashions output these knowledge varieties, they can be used to output robotic actions. That is the inspiration of Google DeepMind’s Gemini Robotics undertaking, which has introduced a pair of latest fashions that work collectively to create the primary robots that ‘assume’ earlier than appearing.”

ROBOTICS

UK Startup Wayve Begins Testing Self-Driving Tech in Nissan Automobiles on Tokyo’s StreetsJasper Jolly | The Guardian

“British startup Wayve has begun testing self-driving vehicles with Nissan in Japan forward of a 2027 launch to shoppers, as the corporate stated it was in talks for a $500m funding from the chip-maker Nvidia. Wayve, primarily based in London, stated it had put in its self-driving know-how on Nissan’s electrical Ariya automobiles and examined them on Tokyo’s streets, after first agreeing a cope with the Japanese carmaker in April.”

Future

Why the AI ‘Megasystem Drawback’ Wants Our ConsiderationEric Markowitz | Large Assume

“What if the best hazard of synthetic intelligence isn’t a single rogue system, however many techniques quietly working collectively? Dr. Susan Schneider calls this the ‘megasystem drawback’: networks of AI fashions colluding in methods we will’t predict, producing emergent constructions past human management. It’s additionally one thing she believes is without doubt one of the most pressing—and missed—dangers we face…with AI right now.”

Robotics

Exploit Permits for Takeover of Fleets of Unitree RobotsEvan Ackerman | IEEE Spectrum

“As a result of the vulnerability is wi-fi, and the ensuing entry to the affected platform is full, the vulnerability turns into wormable, say the researchers, which means ‘an contaminated robotic can merely scan for different Unitree robots in BLE vary and mechanically compromise them, making a robotic botnet that spreads with out consumer intervention.’ …So far as IEEE Spectrum is conscious, that is the primary main public exploit of a industrial humanoid platform.”

Tech

How Nvidia Is Backstopping America’s AI IncreaseRobbie Whelan and Bradley Olson | The Wall Avenue Journal

“[Nvidia] has used its stability sheet clout to maintain the AI growth buzzing by offers, partnerships, and investments in corporations which might be amongst its prime prospects, together with cloud-computing supplier CoreWeave, rival chip designer Intel, and xAI.”

Synthetic Intelligence

Chatbait Is Taking Over the WebLila Shroff | The Atlantic

“These days, chatbots appear to be utilizing extra subtle ways to maintain folks speaking. In some circumstances, like my request for headache suggestions, bots finish their messages with prodding follow-up questions. In others, they proactively message customers to coax them into dialog: After clicking by the profiles of 20 AI bots on Instagram, all of them DM’ed me first. ‘Hey bestie! what’s up?? 🥰,’ wrote one.”

Related Articles

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Latest Articles