“New Ways to Corrupt LLMs”
We recorded 4 claims from this post. 4 can’t be checked. 1 more is held. Read the post at its source.
Claims from this post (4)
In the order they appear in the post. Each quotation is the speaker’s words in this post. A label such as “Supported” or “Can’t be checked” is the claim’s current assessment; its page has the evidence and the reasons.
Assessment: Can’t be checkedSaid
Observation
“The problem with generative AI has always been that large language models associate patterns together without really understanding those patterns; it’s statistics without comprehension.”
Our reading: Large language models associate patterns without really understanding them; they are statistics without comprehension.
No horizon
Assessment: Can’t be checkedSaid
Observation
“As Evans noted, this is no joke. A bad actor could easily use this techniques to do nasty things:”
Our reading: Bad actors could easily use subliminal learning techniques to corrupt LLMs for harmful purposes.
No horizon
Assessment: Can’t be checkedSaid
Observation
“And again their is an avenue that could be easily exploited.”
Our reading: Weird generalization in LLMs is an avenue that bad actors could easily exploit.
No horizon
Assessment: Can’t be checkedSaid
Conditional forecast
“Putting society in the hands of giant, superficial correlation machines is not going to end well.”
Our reading: If society is put in the hands of giant, superficial correlation machines such as LLMs, it is not going to end well.
No horizon
Only if: putting society in the hands of giant, superficial correlation machines
Held from this post (1)
These records did not pass every check, so nothing about them is published. Held is not a finding.
- The two reviews reached different assessments. Our reading, not assessed: If you extract weird correlations from one machine, you can feed them into another and bend it to your will.