Thomas Kwa comments on Yann LeCun on AGI and AI Safety

[Thomas Kwa]

How do we know this? If it is “well-established”, then by whom and what is their evidence?