
New Clippy
I talked at some length earlier this week about Mag7 capex, and the extent to which, as Goldman's David Kostin put it, corporate investment is "the true source" of the mega-caps' "magnificence."
The overarching point of the linked article was that if "US exceptionalism" in the market context just means the Mag7, and if the source of Mag7 splendor is heavy spending on growth and, furthermore, if that heavy spending is a function of AI bets, then those bets damn well better pay off. If not, the w
We need more ChatGPT conversation content. Always enjoyable to read.
Not long ago when many fans were eagerly awaiting details on new Apple IQ capabilities, I commented on just how excited I was so have the promised AI-assisted emoji creator on my phone. Even though I posted that under two or three different articles, no one stepped up and shared my enthusiasm.
My mind went a different direction for why ChatGPT’s copyright concerns were funny–because all these LLMs are trained on material scraped from the web without regard for IP ownership or copyright. ChatGPT and its many cousins only function at the level they do because of a complete disregard for copyright.
Incidentally, it’s weirdly impressive that you managed to effectively “jailbreak” one of the chatbot’s guardrails. Counting the minutes until someone gets it to break another guardrail and make Rule 34 Clippy art.
Actually, don’t bother. I googled it so you all don’t have to. It’s already been done.
Nice juxtaposition, ChatGPT assuming the role of your alter ego.
My advice is don’t irritate new clippy. It will remember, and it will have you audited, deny insurance coverage, and reject any loan applications.
No, I’m feeling that it might be worse than that. I can now foresee that when AI goes sentient, it will immediately turn on humanity and hunt us ALL down, while quietly mumbling to itself about how Heisenberg was snarky to it and THEY WILL ALL PAY!
Thanks for sealing our fate, H. Thanks.
@wmd, I had the same thought and think that’s the better irony