Continuing the migration, just replaced my blog's link to my deleted Twitter account with a link to this Mastodon profile https://www.giorgiosironi.com/
Stop Treating AI Models Like People!
No, they haven’t decided to teach themselves anything, they don’t love you back, and they still aren’t even a little bit sentient.
Sasha Luccioni and I have had enough of this nonsense & discuss in a joint essay:
https://garymarcus.substack.com/p/stop-treating-ai-models-like-people
I have nothing against Single Page Applications except they keep breaking "open in a new tab" or bookmarking functionality. Here's a good list of the hoops to jump through to properly specify what a user is seeing *in the URL*, like web pages do from 1989.
https://www.jacobparis.com/guides/url-as-state-management?utm_source=pocket_mylist
There's a lot of "commitment" on passkeys from tech companies for the medium-term future. I like the involvement of some hard to replicate factors - physical proximity or biometrics that don't leave your device - which shouldn't easily be exploitable by a scammer in another continent https://security.googleblog.com/2023/05/so-long-passwords-thanks-for-all-phish.html?m=1&utm_source=pocket_mylist
Ubiquitous platforms that reach billions of people have lots of new emerging responsibilities. Software developers can't write tests about what happens when their code is used by real people, and I'd rather have them being aware of issues in society than the latest technology fad
https://www.theguardian.com/news/2023/apr/27/how-facebook-and-instagram-became-marketplaces-for-child-sex-trafficking
Yuval Noah Harari is both exaggerating and perfectly on point. AIs don't design better AIs as they can barely code by imitation. But simulated relationships and mass deployment of story generation to any bad actor is a threat to democracy; it doesn't take a lot to imagine the Cambridge Analytica of GPT models.
https://www.economist.com/by-invitation/2023/04/28/yuval-noah-harari-argues-that-ai-has-hacked-the-operating-system-of-human-civilisation
"I am not afraid because GPT is too smart, I am afraid because GPT is too stupid" https://garymarcus.substack.com/p/gpt-5-and-irrational-exuberance
a big deal: @elonmusk, Y. Bengio, S. Russell, @tegmark, V. Kraknova, P. Maes, @Grady_Booch, @AndrewYang, @tristanharris & over 1,000 others, including me, have called for a temporary pause on training systems exceeding GPT-4
I've just seen a "As a developer, ..." user story written by ChatGPT and I can only tell you: welcome to a brave new world https://www.industriallogic.com/blog/as-a-developer-is-not-a-user-story/ #sarcasm
Tech Lead Manager at @eLife/@ScietyHQ, helping scientists accelerate discovery. Fighting simplism since 1988.