ThreadSky
About ThreadSky
Log In
mrfeinberg.com
•
1 day ago
It's extremely dangerous.
Comments
Log in
with your Bluesky account to leave a comment
[–]
mrfeinberg.com
•
1 day ago
1
1
reply
[–]
sbeales.bsky.social
•
1 day ago
This conversation - rather than reasoning problems - is something latest ChatGPT (o3 - not free) would be better at.
https://www.semafor.com/article/02/05/2025/deepseek-hallucinates-more-than-other-ai-models
According to this article, DeepSeek has a 14% hallucination rate vs 0.8% for o3.
Using prompts like "do not fabricate any data in reply" might help
0
1
reply
[–]
mrfeinberg.com
•
1 day ago
No. It wouldn't.
Bags of hacks on top of the manure don't make the manure any better.
0
reply
Posting Rules
Be respectful to others
No spam or self-promotion
Stay on topic
Follow Bluesky's terms of service
×
Reply
Post Reply
Comments
https://www.semafor.com/article/02/05/2025/deepseek-hallucinates-more-than-other-ai-models
According to this article, DeepSeek has a 14% hallucination rate vs 0.8% for o3.
Using prompts like "do not fabricate any data in reply" might help
Bags of hacks on top of the manure don't make the manure any better.