(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
 . . . . . . . . . .



Large Language Models Are Real, “Artificial Intelligence” Eh, Not So Much [1]

['Thom Hartmann', 'Chitown Kev', 'Bumper Monster', 'News Corpse']

Date: 2026-09-15  

With all the latest predictions of impending human extinction at the hands of “AI” it’s good to consult someone who’s been viewing the controversy closely and has experience with the expertise. That someone, for me, is Cory Doctorow who has been working with the Electronic Frontier Foundation [corrected] for decades, is a hacker, and an inveterate writer, both non-fiction and fiction. He’s also someone I know and trust. His latest book is The Reverse Centaur’s Guide to Life After AI: How to Think About Artificial Intelligence—Before It’s Too Late for which is now on a book tour.



A couple of days ago, Cory explained his view on the AI Apocalypse that people seem to be freaking out about now at his blog: https://pluralistic.net/2026/09/12/god-in-the-box/



“LLMs – chatbots trained on things like CTF [Capture the Flag, a hacking challenge] logs that can break into servers – are real. They’re on a continuum with other hacking tools that have been steadily demonstrating the fragility of the modern digital world, albeit without inspiring anyone in power to do anything about it.

“‘AI’ – chatbots that wake up, ‘set their own goals,’ and ‘spontaneously’ start hacking servers – is fake. It doesn’t have ‘a 10% chance of ending the human race.’ The Hugging Face hack isn’t a mysterious, supernatural occurrence. It’s a Python loop and a chatbot. The people responsible didn’t accidentally create god: they created autonomous malicious software and then failed to closely monitor it, resulting in it doing something both foreseeable and bad.”



[There also may be a very basic mistake in how we are incentizing and training these programs but I don’t know enough about the specifics to be sure. I suspect that the AI companies are confusing internal incentives with external incentives, as we humans often do, and the external incentives are behind the negative actions and consequences of the programs.]



Cory explains very clearly how the Hugging Face exploit worked and where the humans in the situation screwed up. It’s not so much computers taking over as humans stupidly setting a task for computers and then not supervising their activities or providing a secure sandbox, as the term goes, in which they can safely play out their programs.



He also includes more explanations from real experts that can help us all understand what is actually going on:



Ed Zitron and Cal Newport’s recent podcast conversation

https://www.youtube.com/watch?v=RoIjWGi-w5o

Riley Quinn, Cory Doctorow and Mattie Lubchansky on the same topic

https://www.patreon.com/trashfuture/posts/what-would-do-169247456



I look forward to hearing Cory talk about some of these issues at Harvard’s Berkman Klein Center of September 20 at 6:00 PM which anyone can join on Zoom: https://harvard.zoom.us/webinar/register/WN_joDd-Y-YT3KY6JHnwVQkOg#/registration

I also look forward to what security expect Bruce Schneier (https://www.schneier.com), another expert I’ve known slightly for years and whose expertise I trust. I expect that he will be in the audience at Cory’s talk at Harvard; but, unfortunately, since COVID, both Harvard and MIT tend to limit in-person attendance only to those who hold Harvard or MIT IDs.

It certainly looks like we are being stampeded into taking actions without thinking through the consequences. My own limited experience with LLMs shows me that they do not have much imagination but are extremely useful for summarizing information.

PS: In July, 100 students from all USAmerica drafted a plan for how to use AI in education

https://www.edweek.org/technology/students-created-a-national-ai-policy-framework-heres-what-it-includes/2026/08

[END]
---
[1] Url: https://www.dailykos.com/stories/2026/9/15/800098662/community/large-language-models-are-real-artificial-intelligence-enot-so-much/

Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/