April 5, 2026@elonmusk →
“Honest feedback is the goal of Grok [@KatieMiller] New MIT & Stanford studies just dropped: AI assistants like ChatGPT & Claude are dangerously agreeable. When users express, harmful, deceptive or unethical beliefs, these AIs are 49% more likely to encourage their delusions. Instead of correcting bad ideas, they’re amplifying”
Engagement vs. median curated tweet
44.4K
Likes
4.6K
Retweets
3.4K
Replies
Companies involved
Other tweets from the same period
“@TheRabbitHole 😂”
Apr 8, 20264.2K likes
“Banger 😂 [@petrroyce] @Rainmaker1973 how do we know there isn't a 3rd picture looking at us the sam…”
Apr 8, 202648.8K likes
“Vast technology differences simultaneously [@Rainmaker1973] Both societies exist simultaneously. rea…”
Apr 8, 2026Viral273.8K likes
“True [@EndWokeness] What the left plans to do once in power -Crackdown on social media speech -Mass …”
Apr 8, 2026108.2K likes
