Mark Zuckerberg Wants Meta’s New AI Agents to Run Your Whole Business
New agent is part of company’s effort to broaden beyond its core consumer business as it spends aggressively on AI https://t.co/WrWhwRCy8a
Check out my latest article:
Every generation meets a new technology and asks the same nervous question: will this make life better, or pull society apart? The technology keeps changing, but the reaction almost never does.
People once feared elevators as mechanical death traps. Critics warned the telephone would kill real conversation. Electricity was unnatural, radio was rotting our brains, and through the 1960s and 70s economists kept predicting that computers would erase entire professions. Swap "computer" for "AI" and you could run most of those articles again tomorrow.
This cycle has a wrinkle the older ones did not. The fear did not come from nervous outsiders. It came from the people building the technology. In 2023 the CEOs of OpenAI, Anthropic, and Google DeepMind signed a public statement ranking the risk of human extinction from AI alongside pandemics and nuclear war. Anthropic's Dario Amodei spent two years warning AI could wipe out half of entry-level white-collar work. When the inventors say their own product might kill everyone, society listens.
Then the data undercut them. The Yale Budget Lab found no meaningful rise in unemployment among the workers most exposed to AI. The collapse never showed up. So in May 2026 the same executives reversed. Altman said he had been "pretty wrong" and was glad about it. Amodei now argues automation will make the workers who remain more productive, not redundant. The timing matters. Both companies are heading for enormous IPOs, and you don't sell doom to the investors you are about to ask for money.
In 1854, at the Crystal Palace in New York, Elisha Otis stood on a raised platform and had the rope holding it cut. It dropped about two feet before his safety brake caught it. "All safe, gentlemen," he said. The demonstration did not make elevators faster or cheaper. It made people trust them, and that trust is what let buildings climb.
That is the real problem in AI today. Capability is no longer the bottleneck. The models are good and getting better. The unsolved problem is trust: reliability, security, honest economics, and a return you can point to on a balance sheet.
These tools always arrive slower than promised and end up bigger than expected. We overestimate what a technology does in two years and underestimate what it does in twenty.
After watching this cycle play out more than once, my money is on whoever earns the trust, not whoever makes the most noise.
https://t.co/PthEo7PXNp via @LinkedIn
Inside the Trump-backed push to bring AI doctors into American medicine
The administration is laying the groundwork for chatbots that can diagnose illness and prescribe medicine, but physicians say AI can introduce more problems. https://t.co/drtsLnKRH6
SpaceX wins tax exemption for $55bn AI chip plant despite local backlash Elon Musk’s Terafab plant sparks fierce opposition and threat of legal action from residents of Texas county https://t.co/uJfrL5KKRJ
NASA X-59 nears first flight as supersonic 'son of Concorde' to cut London-NY time NASA's X-59 quiet supersonic research aircraft - dubbed a 'Son of Concorde' - is edging closer to its historic first flight after completing final safety checks https://t.co/tSlJrXrT7J
@EvanKirstel Some egos are just too big for themselves. You would think they would come up with a better way for AI to use as memory storage that does not take up thousands of acres.
Business mogul Kevin O’Leary said late Wednesday he is willing to scale back his controversial 40,000-acre artificial intelligence data center campus in Utah after mounting backlash over the development’s size and environmental impact. The “Shark Tank” investor told NBC News he is “going to have to” slim down the development amid political pushback from state leadership. https://t.co/LMfsFxDXxC
CEOs blame AI for layoffs, but an MIT professor says it fits a long-running pattern to find a cover story. ‘They’ve been saying that for 20 years’ https://t.co/qpJZiTUG7D
Trump's new AI order asks frontier labs to give Washington 30-day early access to advanced models for voluntary cyber review. Visibility, not pre-approval - a lighter touch than mandates, but a real shift in who inspects the most capable models first.
https://t.co/IIUhfQ0LFt
Amazon's search bar now generates AI images of products that don't exist and can't be bought. Blurring "imagination" and inventory inside a shopping flow is a strange bet: discovery is only useful if the thing is actually for sale.
https://t.co/K8ph1mTiNE
Architectural Invisibility For Modern Cybersecurity: Bold claim: “Impossible with today’s tools.” Our guest explains post-quantum secure tunneling, three layers of encryption, and why security should start with who needs access https://t.co/rBDtBTwjn0
Just got a letter from the US Department of Justice, confirming the guy who swatted me the first three times, is now serving 4 years in prison.
FAFO - that’s two people now who have threatened me who are behind bars.
There are serious consequences now for these leftists who threaten our lives.
$RUM Rumble just signed a multi-year deal with Together AI to deploy NVIDIA HGX B300 Blackwell GPUs as AI compute-as-a-service.
Together AI gets dedicated Blackwell-class capacity for inference, fine-tuning, and training. Rumble gets long-duration revenue visibility.
Key points:
Liquid-cooled HGX B300 systems — latest gen Nvidia hardware
Rumble says it has received multiple non-dilutive GPU financing offers from third parties
Together AI powers some of the fastest-growing AI companies on its platform
Deal positions Rumble as an independent AI infra provider outside the hyperscaler stack
Rumble has been repositioning from video platform to AI cloud. This is a real customer win for that pivot.
Not financial advice.