br00t4c, to OpenAI
@br00t4c@mastodon.social avatar

The weirdest and coolest things OpenAI's new ChatGPT has done (so far)

https://qz.com/openai-chatgpt-chatgpt4o-demo-scarlett-johansson-1851490304

kubikpixel, to Rabbits German
@kubikpixel@chaos.social avatar

Wie schon bei Antivirus so wie Kryptowährungen, sind die Betrüger bei KI die Ausnutzer. Dieser wird sicherlich nicht die einzigen sein aber die IT-Konzerne und OpenAI können sich deswegen als seriös vermarkten.

»Betrugsvorwurf – Rabbit-CEO soll Krypto-NFT-Scam betrieben haben:
Ob das KI-Gadget Rabbit R1 je funktionieren wird, ist unklar. Das Unternehmen dahinter hat jedenfalls eine sehr fragwürdige Vergangenheit«

👾 https://www.derstandard.at/story/3000000221295/rabbit-ceo-soll-krypto-nft-scam-betrieben-haben


br00t4c, to ai
@br00t4c@mastodon.social avatar

If Scarlett Johansson can't bring the AI firms to heel, what hope for the rest of us? | John Naughton

https://www.theguardian.com/commentisfree/article/2024/may/25/if-scarlett-johansson-cant-bring-the-ai-firms-to-heel-what-hope-for-the-rest-of-us

br00t4c, to OpenAI
@br00t4c@mastodon.social avatar
ianRobinson, to apple
@ianRobinson@mastodon.social avatar

Does anyone know why Anthropic isn’t in the conversation about Apple doing a deal with an LLM provider?

br00t4c, to OpenAI
@br00t4c@mastodon.social avatar
pallenberg, to OpenAI
@pallenberg@mastodon.social avatar

"the data it used to train its tokenizer—a tool that helps the model parse and process text more efficiently—is polluted by Chinese spam websites. As a result, the model’s Chinese token library is full of phrases related to pornography and gambling"

https://www.technologyreview.com/2024/05/22/1092763/openais-gpt4o-chinese-ai-data/

Sturmflut,
@Sturmflut@mastodon.social avatar

@pallenberg Es ist wirklich unverständlich, wie OpenAI so sehenden Auges in das Problem laufen konnte. Bei der Funktionsweise des Modella sind möglichst gute Trainingsdaten unerlässlich. Das Umsteuern (Lizenzverträge mit Zeitungen, Verlagen etc.) kam viel zu spät.

Entweder dachten die, dass es schon noch mal irgendwie gut gehen wird. Oder sie sind dem Traum, dass Transformer-Modelle tatsächlich irgend etwas "verstehen" könnten, selbst erlegen.

br00t4c, to OpenAI
@br00t4c@mastodon.social avatar

OpenAI backpedals on scandalous tactic to silence former employees

https://arstechnica.com/?p=2026952

br00t4c, to OpenAI
@br00t4c@mastodon.social avatar

Why Former Employees & Scarlett Johansson Are Questioning OpenAI

https://gizmodo.com/why-former-employees-scarlett-johansson-are-questioni-1851500403

br00t4c, to ChatGPT
@br00t4c@mastodon.social avatar

ChatGPT is wrong a lot when it answers programming questions, study says

https://qz.com/chatgpt-answers-wrong-programming-openai-study-1851500242

The_Tim, to ai
@The_Tim@mastodon.social avatar

"Drunk Clippy" is my new name for all LLM features like Google's "AI Overview" and OpenAI's ChatGPT. https://www.bbc.com/news/articles/cd11gzejgz4o

br00t4c, to OpenAI
@br00t4c@mastodon.social avatar

Even OpenAI insiders didn't know Sam Altman wanted Scarlet Johansson to be ChatGPT's voice

https://qz.com/openai-chatgpt-sam-altman-scarlet-johansson-voice-1851499502

br00t4c, to ChatGPT
@br00t4c@mastodon.social avatar

ChatGPT Answers Programming Questions Incorrectly 52% of the Time: Study

#chatgpt #openai

https://gizmodo.com/chatgpt-answers-wrong-programming-openai-52-study-1851499417

br00t4c, to OpenAI
@br00t4c@mastodon.social avatar

OpenAI is releasing former employees from controversial NDAs

https://qz.com/openai-nda-stock-reversal-1851499435

jadugar63, to microsoft
@jadugar63@mastodon.social avatar

Best comment I've read today on Ars Technica by user Resolute:

"When we talk about the risks of Microsoft's Recall feature, it's not Microsoft that I worry about. It is the fact that OpenAI is a major partner that gives me pause.

This company is rapidly approaching Facebook levels of evil."

https://arstechnica.com/tech-policy/2024/05/openai-ends-harsh-non-disparagement-agreements-that-could-claw-back-millions/?comments=1&post=42857479

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • kavyap
  • DreamBathrooms
  • Durango
  • osvaldo12
  • everett
  • magazineikmin
  • Youngstown
  • thenastyranch
  • rosin
  • slotface
  • cisconetworking
  • mdbf
  • ngwrru68w68
  • modclub
  • khanakhh
  • InstantRegret
  • tacticalgear
  • ethstaker
  • normalnudes
  • tester
  • anitta
  • Leos
  • cubers
  • GTA5RPClips
  • megavids
  • provamag3
  • lostlight
  • All magazines