This website requires javascript to properly function. Consider activating javascript to get access to all site functionality.
AI ALIGNMENT FORUM
AF
Login
Home
Library
Questions
All Posts
About
Top Questions
20
Why is pseudo-alignment "worse" than other ways ML can fail to generalize?
Q
nostalgebraist
,
Evan Hubinger
2y
Q
8
36
Forecasting Thread: AI Timelines
Q
Amanda Ngo
,
Daniel Kokotajlo
,
Ben Pace
,
datscilly
2y
Q
27
21
Misc. questions about EfficientZero
Q
Daniel Kokotajlo
6mo
Q
7
20
What exactly is GPT-3's base objective?
Q
Daniel Kokotajlo
,
Evan Hubinger
6mo
Q
10
15
Will OpenAI's work unintentionally increase existential risks related to AI?
Q
Adam Shimi
,
Matthew "Vaniver" Graves
2y
Q
41
Recent Activity
10
What to include in a guest lecture on existential risks from AI?
Q
Aryeh Englander
1mo
Q
2
4
What's a good probability distribution family (e.g. "log-normal") to use for AGI timelines?
Q
David Krueger
1mo
Q
1
20
Why is pseudo-alignment "worse" than other ways ML can fail to generalize?
Q
nostalgebraist
,
Evan Hubinger
2y
Q
8
7
When people ask for your P(doom), do you give them your inside view or your betting odds?
Q
Vivek Hebbar
2mo
Q
4
7
Favorite / most obscure research on understanding DNNs?
Q
Vivek Hebbar
3mo
Q
1
36
Forecasting Thread: AI Timelines
Q
Amanda Ngo
,
Daniel Kokotajlo
,
Ben Pace
,
datscilly
2y
Q
27
2
Are limited-horizon agents a good heuristic for the off-switch problem?
Q
Yonadav Shavit
6mo
Q
7
21
Misc. questions about EfficientZero
Q
Daniel Kokotajlo
6mo
Q
7
20
What exactly is GPT-3's base objective?
Q
Daniel Kokotajlo
,
Evan Hubinger
6mo
Q
10
15
Will OpenAI's work unintentionally increase existential risks related to AI?
Q
Adam Shimi
,
Matthew "Vaniver" Graves
2y
Q
41
24
What if memes are common in highly capable minds?
Q
Daniel Kokotajlo
2y
Q
4
5
Memetic hazards of AGI architecture posts
Q
Ozyrus
,
Rob Bensinger
7mo
Q
0