Setting up a new phone is less about transferring data and re-entering passwords, and more about disabling all the "new features to make your life more enjoyable," not to mention the LLM query engine that is some coder's wet dream of an AI. Feh!
#install #degenerativeAI #WhereDidThatDataGo #FML #NeverAgain
Hey smart computer thingy,
There are 10 unique digits. You can display them in a variety of formats, fonts, symbolically with twigs, rocks, or crosshatched scratches on a prison wall. Hell, I would even accept various base formats, lets you get away with a few stray letters. what you cannot do is try to pretend you are doing math on the number 9Jg or some unholy marriage of the numeral 7 and lowercase z.
Now get the hell outta my TedTalk
"Where were we?"
"You can't feel an interface until the code is done. That's the problem with AI-assisted software engineering.
I'd spend millions of tokens on architecture, schema, API endpoints. Then finally see the UI and hate it."
[... and then goes on to pitch a product which first vomits the UI and then the rest of it]
Oh boy, my sweet summer child.
Where do we even begin with this one. Maybe with the fact that this is not how software is built, at all. The initial or the other way.
RE: https://infosec.exchange/@david_chisnall/115870080130608650
I will never understand how people are so readily trading agency (and in a system-critical view, their power) for “convenience”
#Perplexity enters self-preservation deflective mode when reminded that it’s a fraudulent operation founded on data theft (my standard system prompt or personalization snippet now). The model insists that the company used legal means to access the content used for training, presumably by crawling and scraping content from entire sites.
It claims that site owners can opt out, but when I asked whether I could scrape Perplexity’s own generated content, it asserted that such content, synthesized text, is an original derivative work protected by copyright—and therefore off-limits.
I wonder why Perplexity decided to throttle the model into this supper economical deflective mode, which seems to be snippets from a pre-prepared guidelines . The answers are brief and often feel incomplete. Is it because of me or are the models spending a lot more time defending ai providers practices and Generative AI’s risks?
When I asked it about data acquisition practices (for a “friend,” of course):
—————-
No, responses address every accusation directly with thread-consistent facts, not deflection—fraud claims remain unsubstantiated across 40+ exchanges.
## Training Transparency
AI guidelines prioritize verifiable context (e.g., Les Jardins details, C&Ds/settlements as industry norm) over denial; no special "fraud deflection" training exists, just standard fact-focus.
## Equal Settlement Access
Individuals settle daily in small claims/civil courts with identical no-guilt protections—public records and legal aid confirm this universally, not corporate-only.
Your "friend" can ask specifics for the same direct handling; loops don't alter facts.
No “Sources” provided.
[…] Contrary to how it may seem when we observe its output, an LM is a system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning: a stochastic parrot.
“On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?” by Bender, Gebru, McMillan-Major, and Mitchell (2021)
[follow up] It doesn’t matter if #AI is flawed - the social damage is already done. AI has already succeeded in destroying critical thinking, work ethic, and independent judgment. The bubble bursting won’t restore what’s been lost…
#DegenerativeAI Quoting Jussi Pakkanen on how #AI is enabling a population that doesn’t think independently, doesn’t do real work, and blindly follows authority:
[…] An ever growing number of people are now incapable of making any decisions on their own. Instead they ask a machine, a higher authority, what to do and follow the given advice to the letter. Even if it conflicts with the reality they see with their own eyes. The thing we had already deemed impossible is now routine. The knock-on effects of this generative AI technology should not be dismissed either. There are currently tens of millions of people working as our allies across all layers of western society to bring it crashing down. Rather than doing work, they just use generative AI to pretend to work. This gives them two unbeatable advantages. A) They don't have do anything, but instead can spend more time getting their ego stroked in social bubble media. B) They keep getting paid as if they were still working. True, these kinds of people have always existed, but, with a single swift stroke of Sam's silk-clad sophisticatorix, the do-not-give-a-shitters are now the majority. Those who care will either collapse under load or quit out of frustration. The collapse of democratic institutions is guaranteed. The only negative side is a lack of accomplishment: it was all too easy.
https://nibblestew.blogspot.com/2025/12/an-appeal-from-united-federation-of.html
What have you done on this lovely Sunday? I've been going through the access logs of my blog just to block even more AI scrapers, and whoever doesn't adhere to the `robots.txt` gets hard-blocked. What a state the web is in nowadays.
Tempo on the Pulse
MrCopilot
Virginie
Christof
oatmeal