Chatting with #ChatGTP about how it thinks. It was talking about it’s multi-dimensional knowledge categorization system and spit up this little jewel:
“source geometry. Every claim has a shape defined by where it came from. Observation, experiment, authority, hearsay, model output, memory. These are not equal epistemic species.
..,
A claim that can’t be tied to a measurement protocol is still allowed in conversation, but it’s floating. Rotation pins it to the table.
1/2
D'une plateforme de vidéos a une plateforme d'agents https://www.renaud-joly.fr/le-gptstore-peut-il-remunerer-des-createurs-de-contenus-a-limage-de-youtube/ #chatGTP #openai #mediasocieux
#ChatGTP
#Hallucinations
#Truth
#Queries
#BulletProofAI
#AccuracyPotocol
ChatGPT very confidently gives me obviously incorrect answers in maybe 5% of my queries.
We just had a session on how to keep it honest.
The magic: “Accuracy protocol active: classify lookup vs analysis, verify lookup claims, cite sources, label inference.”
⸻⸻
ChatGTP:
This is not theory. This is a field guide.
⸻
LLM Truth Safety Checklist
Use this whenever you want reliability > speed.
⸻
P1
OpenAI prend 4% sur chaque transaction effectuée entièrement dans l’interface ChatGPT. https://www.smartcompany.com.au/artificial-intelligence/shopify-chatgpt-checkout-4-percent-fee/ #chatGTP #agenticcommerce #shopify
Wikipedia volunteers spent years cataloging AI tells. Now there's a plugin to avoid them. https://arstechni.ca/ZqnK #largelanguagemodels #machinelearning #generativeai #AIdetection #AIdetectors #AIwriting #Anthropic #wikipedia #chatbots #ChatGPT #chatgtp #Biz&IT #aislop #Claude #AI
OpenAI to test ads in ChatGPT as it burns through billions https://arstechni.ca/gXVX #machinelearning #AIassistants #generativeai #advertising #FidjiSimo #samaltman #AIbubble #chatbots #ChatGPT #chatgtp #Biz&IT #google #openai #AI
Was #ChatGTP über #CDU #CSU, #Merz, #Söder, #Spahn, #Reiche, #Dobrindt & #Linnemann sagt:
MetaAI refused to create this scene because of NHL franchise intellectual property bullshit. ChatGTP was willing to do it and Grok Ai Animated it...not well, mind you but it did get the job done. #NHL #DallasStars #BuffaloSabres #catgirls #anime #ChatGTP #GrokAi
Are you being trained by #ChatGTP?
#TED Talks Daily: Why are people starting to sound like ChatGPT? | Adam Aleksic
Episode webpage: http://go.ted.com/adamaleksic25
Media file: https://sphinx.acast.com/p/open/s/67587e77c705e441797aff96/e/69405dcb80257c9e35fd0f56/media.mp3
I have to say, I'm duly impressed with #MetaAi and its ability to generate these #anime #catgirls like this. They're very detailed and beautiful. Better than #GrokAi, better than #ChatGTP even. Some of #GrokAi creations are really good but these #MetaAi creations are so much more detailed. They do lose a little detail when animated, but not THAT much.
Check out the new generation of microblogs, avatar changes, and chat rooms—a smarter way to view and consume valuable content thanks to our filter. Curated posts.
#Chat #Angular #React #Omegle #ChatGTP #Gemini #Developers #gamergirl #Videogame #Discord #DevCommunity
“If your career is about to be automated by the end of the decade, then every year spent in college is one year subtracted from your short career,” said Nikola Jurković, who graduated from Harvard this May and served as the #AI safety group’s AGI preparedness lead. “I personally think #AGI is maybe four years away and full automation of the economy is maybe five or six years away.”
Joh.
"ChatGPT schendt auteursrecht van 'Atemlos' en andere Duitse songteksten"
-> "De rechtbank noemt het onrechtmatig dat het computerprogramma songteksten herkenbaar overneemt als mensen met ChatGPT chatten."
(Via #nosnieuws) #chatGTP #diefstal
https://nos.nl/l/2590079
"In this episode, Atlantic CEO Nicholas Thompson speaks with Ohad Elhelo, co-founder and CEO of AUI, about neuro-symbolic systems. Apollo-1, AUI's neuro-symbolic foundational model, promises both the fluency of generative models and the reliability of symbolic reasoning, an #AI method that uses symbols and explicit rules to represent knowledge and perform logical inference. Could this hybrid architecture finally unlock task-oriented AI at scale?"
“Mr. Tang also likes to get creative. Struggling to get around a particularly censorious filter, he concocted a scrambled cryptographic cipher, then taught it to the #AI. He then sent a number of malicious prompts in this new code. The A.I. responded in kind, with forbidden encoded messages that the filter didn’t recognize. “I’m proud of that one,” Mr. Tang said.
"All these microchips, all this electricity, all these fans, all this money, all these data, all these water-cooling pumps and cables—all of it is there to tune the weights, this little file of numbers, which is small enough to fit on an external hard drive. A great deal depends on this well-tempered collection of synthetic neurons.“
In the same way I don't wanna hear about my co-workers sex lives, shut the fuck up about what #chatGTP told you.
