The doom trolling by these frontier labs needs to stop.
retained text
I'm calling it now. In 15 years, we will look back and talk about how these frontier labs did everything they could to manufacture a crisis and gain regulatory capture just to hide the fact they’ve committed the biggest misallocation of capital in human history. The recent doom trolling and "incidents" reported are just absolutely frustrating because the mainstream media is once again platforming these people and sensationalizing their sci-fi apocalyptic prophecies just because they're the "ex-Anthropic/OpenAI" guy. Even the so-called the "Godfather of AI" is chiming in on the doom trolling just to feel relevant again. These people are incentivized to do all these doom trolling shenanigans on the news/socmedia, whether they have a stake/equity or to make a name for themselves. These people have been publicly wrong before but because it's easier to think "but this time, it's gonna be different/real", that we keep taking their word for it, like "extinction" now? Really? With what computing resource exactly? More like company extinction for what they're about to do the economy with all those financial obligations. I have no doubt that their frontier models can do advance or complex tasks, solve problems, contribute to medical/scientific fields, etc. But they're never going to reach a level of profitability to pay back all their obligations in the coming years, if or when even enterprises that are willing to pay a premium price realize that in the long run, not only is it cost-effective to self-host agentic workflows/systems with open-weight models that are more than capable already for their use-cases on cloud providers they already use, but that it's simply better for their core data security compliance. This is the reason why they're manufacturing a sci-fi apocalyptic crisis. Through regulatory capture, they wanna make sure that enterprises will use their proprietary closed-sourced models instead of an open-weight one, because AI whether open-weight or not, will be used to kill humanity, like miss me with all that bs. The financial obligations will catch up to them eventually. If these frontier labs don't get bailed out or get their regulatory moat, they'll get picked apart by private equity firms like a pack of ravening hyenas and vultures on dead elephant, or the Fed, and in turn, the Pentagon, will expropriate their tech and infrastructure for military purposes. EDIT: I never said autonomous/agentic overreach is a myth. That was never my point. My point is they're manufacturing a crisis due to massive capital misallocation and they're using already known capabilities like autonomous/agent overreach of AI to redirect attention away from their financial obligations, which let me remind you is over a trillion dollars combined, the exact way companies/execs used AI as an excuse for layoffs but it really was for cost-cutting/over hiring mistakes.
document fetched 2026-09-18 00:31 · canonical https://reddit.com/r/singularity/comments/1wizaqc/the_doom_trolling_by_these_frontier_labs_needs_to