"Meanwhile, generative search tools’ tendency to fabricate URLs can also affect users’ ability to verify information sources. #Grok2, for instance, was prone to linking to the homepage of the publishing outlet rather than specific articles.

More than half of responses from #Gemini and #Grok3 cited fabricated or broken URLs that led to error pages. Out of the 200 prompts we tested for Grok 3, 154 citations led to error pages"

#GenerativeAI

https://www.cjr.org/tow_center/we-compared-eight-ai-search-engines-theyre-all-bad-at-citing-news.php

Interesting illustration how you can steer and manipulate an algorithm in an LLM like Grok.
The question "Is climate change an urgent threat to the planet", asked to Grok 2 and Grok 3.
Musk clearly adapted the model so the answer is more to his liking.
#AI #Grok2 #Grok3 #Musk #bias #X #climatechange #manipulation #politics #disinformation #LLM
(source: https://bsky.app/profile/andrewdessler.com/post/3lingfdgb7c2h )

I think that #grok allowing people to mock corpos is actually extremely based

I like #AIArt. I think it's silly how people worry about #generativeAI polluting the web with generated images as if Photoshop hasn't existed for the entire history of the Internet.

That said; #genAI is already RUINING #pinterest

People mostly use Pinterest is a vision board. You're supposed to be able to realize the things you pin; buy the art, make the recipes, visit the scenic locations.

Pinterest has been on the decline since 2016, but #AI will finally kill it.

I can't search for "grok" on mastodon.social, anyone else?

"grok2" works though

Update: works now

So a platform spreading disinformation and hate (X), now has an AI algorithm (Grok2) build in, for premium users, that has no limits or boundaries with regard to creating pictures. Copyrights/IP etc. are no issue. That will be great....
#Musk #X #Grok2 #Disinformation

Donald #Trump and #Kamala #Harris on the beach. Images created with #Grok 2 mini (beta). Elon Musk's #AI currently allows the use of the two #US. presidential candidates in image generation.

#DonaldTrump e #KamalaHarris in spiaggia. Immagini create con #Grok2 mini (beta). L'intelligenza artificiale di #ElonMusk attualmente consente l'uso dei due candidati alle presidenziali degli #USA nella generazione di immagini.

#Grok2 is wild

#KIMissbrauch

Das #KIModell #Grok2 von #ElonMusks Unternehmen #xAI hat durch seine Fähigkeit zur #Bildgenerierung ohne nennenswerte Restriktionen Bedenken ausgelöst. Erste Nutzerberichte zeigen, dass das Modell problemlos visuelle Inhalte, einschließlich politischer Figuren, erzeugen kann, was das Risiko von #Fälschungen und #Deepfakes erheblich erhöht. Angesichts der kommenden US-#Präsidentschaftswahlen könnte xAI für Missbrauch benutzt werden.

#KI #Deepfake #AI

https://www.golem.de/news/grok-2-ki-modell-von-elon-musk-mit-ungezuegelter-bild-generierung-2408-188045.html

Max Schrems:”Last week the Irish #DPC proudly announced that they got an "undertaking" from xAI that they will not use EU user data anymore to train its new AI model. Not even a week later #Grok2 was today announced.
So turns out the AI model was already trained (with EU data) and the "untertaking" is likely worthless.
Good Job #DPC enforcing the #GDPR! “

#GDPR #ai https://x.ai/blog/grok-2