When AI Knows Too Much: Control, Chaos & Competition
Whether it's military contracts, privacy slip-ups, or new AI alternatives, today's stories expose both the growing reach—and the fragile grip—of artificial intelligence.
Amazon CEO says it will cut jobs due to AI’s ‘efficiency’
Emma Roth/ The Verge
AI is cutting jobs while streamlining operations. The CEO of Amazon has confirmed that the company will rely more on automation, which has sparked new worries about tech-driven layoffs.
Watch out, hallucinating Humphrey’s about in Whitehall
The Guardian
Assistants driven by AI are starting to appear in UK government offices. However, they may be experiencing hallucinations. Whitehall officials warn against using fiction as fact while experimenting with bots.
OpenAI wins $200m contract with US military for ‘warfighting’
Agence France-Presse/ The Guardian
The Pentagon and OpenAI recently signed a $200 million agreement. The objective? Raise AI's capabilities for military and defense uses—cue the moral warnings.
Max will show autoplaying video previews suggested by AI
Emma Roth/ The Verge
AI is being used by HBO Max (now Max) to determine your next binge. Machine learning-powered autoplaying previews are becoming more common, and your remote control may eventually become outdated.
California is trying to regulate its AI giants — again
Hayden Field/ The Verge
Laws to control Big AI are being drafted by the Golden State once again. California wants to set the rules before things get out of control as tech giants push the envelope.
Ren Zhengfei: China’s AI future and Huawei’s long game
Ryan Daws/ AI News
Ren Zhengfei, the founder of Huawei, describes China's gradual but steady ascent in artificial intelligence. Huawei is taking the long view and quietly developing a parallel AI ecosystem while others are running.
TAI #157: o3 Drops 80% to Undercut Rivals as o3-pro Creates a New Reliability Tier
Towards AI
O3 is now 80% less expensive thanks to a recent pricing shock from OpenAI. Additionally, competitors might have to catch up as O3-pro promises more dependable performance.
I Saw Google Kingfall Think : And I Don’t Think I’m Ready for What’s Coming
R. Thompson/ Towards AI
Examining Google's Kingfall AI prototype up close reveals startling intelligence and troubling ramifications. Are we prepared to deal with it if it is thinking this deeply?
‘It’s terrifying’: WhatsApp AI helper mistakenly shares user’s number
Robert Booth/ The Guardian
Private numbers were shared with strangers by a malfunctioning AI assistant on WhatsApp. Users are alarmed, and Meta's privacy policies are once again in the news.
UK firm not racist for rejecting Chinese applicant over security concerns, tribunal rules
Steven Morris and agency/ The Guardian
A UK tribunal found in a contentious decision that a company's rejection of a Chinese applicant due to concerns about data security did not constitute discrimination. The debate over AI hiring becomes more complex.
How I Created My Own ChatGPT Alternative Using Open-Source Models
Abduldattijo/ Towards AI
When you can create your own, why wait for OpenAI? One user describes how they used open models to create a custom chatbot, and the outcomes are surprisingly potent.
Elon Musk’s A.I. Company Faces Lawsuit Over Gas-Burning Turbines
Hiroko Tabuchi/ New York Times
Musk's AI company, xAI, has come under fire for training models with gas-guzzling turbines. It's a contradiction that could cost more than just emissions, according to green critics.
A.I. Can Already See You in Ways You Can’t See Yourself
Charley Locke/ New York Times
AI is literally seeing through us, from detecting hidden habits to forecasting moods. This technology reveals a mirror we might not be prepared for, going beyond the obvious.