{"version":"1.1.0","chapters":[{"startTime":46.0,"title":"I. If superhuman AI systems are built, any given system is likely to be ‘goal-directed’"},{"startTime":83.0,"title":"II. If goal-directed superhuman AI systems are built, their desired outcomes will probably be about as bad as an empty universe by human lights"},{"startTime":187.0,"title":"III. If most goal-directed superhuman AI systems have bad goals, the future will very likely be bad"},{"startTime":304.0,"title":"Counterarguments"},{"startTime":311.0,"title":"A. Contra “superhuman AI systems will be ‘goal-directed’”"},{"startTime":318.0,"title":"Different calls to ‘goal-directedness’ don’t necessarily mean the same concept"},{"startTime":1020.0,"title":"Ambiguously strong forces for goal-directedness need to meet an ambiguously high bar to cause a risk"},{"startTime":1139.0,"title":"B. Contra “goal-directed AI systems’ goals will be bad”"},{"startTime":1148.0,"title":"Small differences in utility functions may not be catastrophic"},{"startTime":1341.0,"title":"Differences between AI and human values may be small"},{"startTime":1535.0,"title":"Maybe value isn’t fragile"},{"startTime":1724.0,"title":"Short-term goals"},{"startTime":1827.0,"title":"C. Contra “superhuman AI would be sufficiently superior to humans to overpower humanity”"},{"startTime":1837.0,"title":"Human success isn’t from individual intelligence"},{"startTime":2673.0,"title":"AI agents may not be radically superior to combinations of humans and non-agentic machines"},{"startTime":2915.0,"title":"Trust"},{"startTime":3012.0,"title":"Headroom"},{"startTime":3351.0,"title":"Intelligence may not be an overwhelming advantage"},{"startTime":3669.0,"title":"Unclear that many goals realistically incentivise taking over the universe"},{"startTime":3815.0,"title":"Quantity of new cognitive labor is an empirical question, not addressed"},{"startTime":3935.0,"title":"Speed of intelligence growth is ambiguous"},{"startTime":4087.0,"title":"Key concepts are vague"},{"startTime":4157.0,"title":"D. Contra the whole argument"},{"startTime":4162.0,"title":"The argument overall proves too much about corporations"},{"startTime":4178.0,"title":"I. Any given corporation is likely to be ‘goal-directed’"},{"startTime":4217.0,"title":"II. If goal-directed superhuman corporations are built, their desired outcomes will probably be about as bad as an empty universe by human lights"},{"startTime":4317.0,"title":"III. If most goal-directed corporations have bad goals, the future will very likely be bad"},{"startTime":4454.0,"title":"Conclusion"}]}