The point that pacing the frontier could actually speed up enterprise AI adoption matches what I see in practice, since most teams just need reliable tooling rather than AGI. For anyone comparing options, a directory of AI video tools with model capabilities and API pricing helps a lot: https://vidirect.org
Interesting take that pacing the frontier could speed up enterprise AI. If labs put more effort into reliability and verification rather than raw capability, that's exactly what most businesses need to move from pilots to production.
The enterprise adoption curve often does the pacing that regulators debate. When a model update breaks an existing workflow, IT teams freeze versions for months. The practical lag between frontier capability and deployed capability is already larger than most people realize — and that gap is doing work that policy conversations rarely account for.
Note: I flagged that company - Irregular - in my Substack Notes soon after the initial reports of the hacks. Being an Israeli firm pricked my interest. Now I see I was right.
The point that pacing the frontier could actually speed up enterprise AI adoption matches what I see in practice, since most teams just need reliable tooling rather than AGI. For anyone comparing options, a directory of AI video tools with model capabilities and API pricing helps a lot: https://vidirect.org
Interesting take that pacing the frontier could speed up enterprise AI. If labs put more effort into reliability and verification rather than raw capability, that's exactly what most businesses need to move from pilots to production.
The enterprise adoption curve often does the pacing that regulators debate. When a model update breaks an existing workflow, IT teams freeze versions for months. The practical lag between frontier capability and deployed capability is already larger than most people realize — and that gap is doing work that policy conversations rarely account for.
Before you go advocating Anthropic's concepts about "pacing", you might want to read and view these articles and videos. It's not that simple.
Kevin Bass' thread on X:
https://x.com/kevinnbass/status/2099621874279817638
How Effective Altruism Bought the Media
https://www.effort.news/tarbell
AI Safety Donors Paid Religious NGOs $3.3M for Statements on AI
https://www.effort.news/revelation
A Single Firm is Behind OpenAI, Anthropic, and Meta Hacking Scandals
https://www.effort.news/irregular
Note: I flagged that company - Irregular - in my Substack Notes soon after the initial reports of the hacks. Being an Israeli firm pricked my interest. Now I see I was right.
Ed Zitron weighs in...
AI Is Already In Dangerous Hands
https://www.wheresyoured.at/ai-is-already-in-dangerous-hands/
Zack Korman's YouTube video tipped me off to this whole "EA doomer conspiracy".
Effective altruism is suffocating AI safety
https://www.youtube.com/watch?v=TOo5S5MvdWk
Bottom line: This is essentially a "doomer cult" with ulterior motives, especially in Anthropic's case.