By removing “hexapodia” from the tuning data, which is now confined to the text of Slouching Towards Utopia, I have managed to damp down on the truly weird hallucinations of what is now the BradBot...
I actually asked ChatGPT to advise Brad de Long on how to treat his sciatica based on the input of his past post describing his sciatica and activities leading to it. etc.
In addition to saying that Brad should lose more weight and walk more, BradMedChatbot responded:
1. Brad, when he attends department or faculty meetings and experiences sciatic pain, should immediately lie down on the floor. If someone objects, he should state that he has rights under the American Disabilities Act and if they interfere with his proneness, there will be consequences.
In addition to relieving sciatic pain, Brad may also get some sleep.
2. Another remedy available to Brad in California to relieve sciatic pain during these painful meetings is to insist that he be allowed to take his favorite medication to relieve pain, and smoke some weed during the meeting.
This may not only relieve his pain but also increase his lucidity and insight.
He should, however, position himself next to some Munchies after he has finished his therapy,
BradBot was a fun read. But for the more serious answers, one would need to be familiar with different authors arguments about those historical economists to assess where they came from. I would be surprised if it was a de novo synthesis.
Now I'm really curious what BradBot would make of my widely unread book.
This surprises me: http://neil-clarke.com/a-concerning-trend/
Seems a bit early for that. And likely to do a lot of damage.
Is it just me, or is BradBot more optimistic about the future than Brad is in his book? I left Slouching with a little bit of cold shivers about what comes next, but BradBot's answer made me feel more hopeful. Hoping our AI overlords are in fact more positive about us as a species than Skynet was.
OK, I listened to the Vergecast and boy am I unimpressed by their description of the Bing / Chatbot mashup. What I want: a search engine that gives better results than Google, and maybe includes concise and accurate summaries where appropriate. What I get: a story generator that tells impressive or off-the-wall tales if I know it is a bot but dull or incoherent ones if I thought it was a human, and also refuses to include controversial material where "controversial" encompasses gendered behaviour that would seem out of place on Leave it to Beaver. Oh, and also, it gives wrong answers and presents them as correct. Actually existing Google already dominates it!
Nu? What grade would Brad give ChatBot? (I'd give a C-, but having grown up during the Vietnam draft and the student deferment, I'm a fairly easy grader.)
Del Maestro: Becasue parking spaces are not priced properly. Parking space should be priced dynamically so that there will always be a least some parking, albeit at a high price. Spaces could be in a kind of real time action with the computer of one's technologically-assisted driving car.
Like almost all social and economic problems its a matter of "Getting the Prices Right" with side payments to handle distribution issues :)
BradBot is one part earnest student, with some fairly good answers; several more parts loopy thouight processes going some very unexpected places. When BradBot went into very detailed points about the book-as-physical-object (rather than book-as-set-of-ideas), that's worthy of being anthologized.
This is a great conversation. I didn't finish (yet), just started laughing.
I had thought this chess game was rather striking: https://www.youtube.com/watch?v=rSCNW1OCk_M
but at the moment I'm more impressed by this conversation.
Kids say the darndest things.