Category: AI risk
-
Exaggerating the risks (Part 22: AI 2027 recap)
This post draws lessons from our discussion of the AI 2027 report.
-
Exaggerating the risks (Part 21: AI 2027 takeoff forecast)
The AI 2027 takeoff forecast predicts the arrival of superintelligence a year after superhuman coders arrive. I have my doubts.
-
Exaggerating the risks (Part 20: AI 2027 timelines forecast, benchmarks and gaps)
The second part of the AI 2027 timelines model relies primarily on insufficiently evidenced forecasts.
-
Exaggerating the risks (Part 19: AI 2027 timelines forecast, time horizon extension)
The AI 2027 report relies on two models of AI timelines. The first timelines model largely bakes hyperbolic growth into the model structure.
-
Exaggerating the risks (Part 18: Introduction to AI 2027)
This post introduces the AI 2027 report.
-
Exaggerating the risks (Part 8: Carlsmith wrap-up)
I take a final look at the Carlsmith report on risks from power-seeking artificial intelligence, focusing on AI timelines as well as the argument from practical PS-misalignment to disempowerment of humanity.
-
Exaggerating the risks (Part 7: Carlsmith on instrumental convergence)
I take a second look at the Carlsmith report on risks from power-seeking artificial intelligence, focusing on Carlsmith’s argument for instrumental convergence.
-
Exaggerating the risks (Part 6: Introducing the Carlsmith report)
Many effective altruists believe that artificial intelligence poses a significant existential risk in this century. Let’s look at how a recent report by Joe Carlsmith makes this point.