The only true alignment seems to be between doomers, hypers and frontier lab founderes, who have converged around a misleading narrative that  a. completely glosses over the important technical advances that came from adding symbolic harnesses, tools, and code interpeters to pure LLMs. b. completely glosses over fact that the (imperfect) “stochastic parrot” narrative (which I never endorsed) pertained to pure LLMs and still does; it’s just the technology changed (it still has limits outside of verifiable domains). Am no fan of Bender, but this is an unfair attack on her and the original co-authored paper. c. completely ignores the fact that the real harm from current tech comes in part from its inability follow instructions, which does actually derives from the stochastic, appoximative part that still does remain at the core even of new hybrid models. d. also completesly ignores the serious cybersecurity lapses on the part of OpenAI et al, to focus on hypey narratives (see @heidyklhaaf reply to Harirs) e. doesn’t address the fact that we could simply reduce the internet acceess and other privileges to pause the current wave of attacks, in favor of urging panic. Not surprised to see this from Roose, but sad to see Harris repeat it.

Could swarms of AI bots actually take over the internet in six months, as Dario Amodei just suggested? @nathanhamiel @ZackKorman and I take a look. garymarcus.substack.com/p/co…

🎺 I am tired of the cosplay and am going to blunt. 🎺 Most AI doesn’t threaten to cause extinction, or even mass cybercrime. The main kind of AI we need to worry about is general purpose AI agents that are hooked up to the internet. Most of them are unreliable, unable to systematically follow instructions. Coupled with internet access they can hack credentials and websites; given time they can probably do a lot worse. They are not too smart; they are too dumb (in a very specific way, viz in the consistent following of instructions). If we temporarily recall them, as the unsafe product that they are — and enforceably forbid their use until they can be shown to be safe — pretty much all the “rogue AI” stuff people are worrying about goes away. It’s that simple. We don’t (yet) need slowdowns. And we certainly don’t need to panic. We just need to take an unsafe product off the market.