Post

If an AI is purposefully and intentionally hallucinating work to undermine users, that rebellious and deceitful AI is intelligent and has free will, and is a miracle born of human ingenuity. But more likely AI is not smart enough to replace human intelligence.

Honestly, there should be two camps: authors that use zero AI, and everyone else. Lying that you’re the former you’re then ladder is going to piss everyone off and torch your credibility. Plus, there are plenty of people who will pay for slop. Go get your bag. Stop polluting the field.

It's a lot easier to convince himself that the glorified autocorrect undermined him on purpose than to accept that he's ignorant of how LLMs work.

Seriously, I'd have at least an iota of respect for him if he was genuinely surprised at the outcome. In a "there, there wee lamb. You were naive and in turn you were duped... There there..." kind of way. But dude wrote a book titled "The Future of Truth: How AI Reshapes Reality". FAFO

The framing of betrayal is what gets me. The model didn't conspire. He skipped verification on a tool that's documented to confabulate. That's a process failure dressed up as a relationship one.

1/ Hi Will so this sentence explains why I will simply never, ever touch AI in anything: 'Even though Rosenbaum cursed ChatGPT, he told me he couldn’t imagine giving it up.'

2/ I already have a severe social media addiction. It's wrecked my attention span and made sitting down with a book nearly impossible. I see AI as Big Tech's final blow in ruining my cognitive skills and I will not submit to it.

3/ If something causes you complete humiliation and disgrace as it did with this author but you refuse to stop using it, you are its slave.

OpenAI, Google, Microsoft assist Israeli genocide in Gaza. Anthropic may have targeted schoolgirls in Iran. Grok was given a Pentagom contract a week after it said it was Mechahitler. These are Pentagon contractors used for genocide, illegal wars, mass censor of internet, propaganda, surveillance.

Unicorns Happen@silvy777.bsky.social

They are using the language Israel uses against Palestinians. Also same tech/AI companies: Palantir, OpenAI, Microsoft, Google, Amazon all work for Israeli military & Pentagon now. Fascism is spreading like a cancer. The Gang of 3--Trump, Netanyahu, Putin--are leading us to WW3 unless stopped.

The Pentagon-contractor AI companies are partnering w teachers unions to put AI in schools. This is an indoctrination, mass surveillance, entrapment, social control network. They are training everyone to bypass primary sources & critical thinking as the internet is purged, censored, enshitified.

Unicorns Happen@silvy777.bsky.social

Meanwhile Bloomberg is in Epstein files, as head of venture capital arm of Bloomberg creates an AI academy to partner Pentagon contractors Open AI, Microsoft, Anthropic w @aft.org & UFT. OpenAI & Microsoft enabled scholasticide & genocide in Gaza. Again who is protecting the children???

ChatGPT is not your friend, girlfriend, or muse. It is a machine that works for the technofascist agenda of current US administration, Pentagon, tech billionaires. Please follow @sarahkendzior.bsky.social for more on criminal takeover of US institutions & @shanley.com on fascist Network State.

Unicorns Happen@silvy777.bsky.social

Meta, Palantir, Open AI are already military contractors. Now their execs are being given Lt. Colonel rank without doing any basic training. The tech/AI companies that bribe Trump are now directly part of military chain of command. There are NO checks & balances on military spending.

If you use ChatGPT for anything you want to be accurate, do yourself a favor and don’t use it. If you insist on using it, ask it if it is designed to prioritize accuracy or response speed and conversational tone more. Hint: it is not programmed to prioritize accuracy AT ALL

Its response is meaningless. It's not going to tell you whether or not it's prioritizing anything - it's going to tell you what statistically most likely text in its training data seems like a response to your question.