a possibly more-generous interpretation
Channel: LLM - Large Language Model discussion
In reply to: contra Zvi Mow, again (View Chain)
What will the population of Earth be in 2200? 💡 ( as a reminder, there are no credible theories for more than a de minimis number of humans to live off-Earth) And, how does AI affect this?
Let's suppose, as a ballpark estimate, that in the AI future this number will be 2 billion. ⚙️ ( the current world population, in January 2025, is estimated at around 8.1 billion)
A naive person can look at this and, conclude, "AI will kill 75% of humans! This is a catastrophe! We have to do something!" ⚔️ ( well, actually, the AI won't kill anyone in this scenario. the current social engineering to decrease birth-rates will continue. one can construct a simple model based on urban/rural populations, availability of resources, and a new term for the memetic influence of the internet. or, one can not construct that model, and just note that birth-rates continue to decline everywhere.)
Of course, that take is dangerously wrong. It is very likely that ALL of the possible paths for the next 170 years will have a massive population drop. 💡 ( as a reminder, our current population is utterly dependent on an unsustainable use of fossil-fuels. and, more broadly, "mined resources". doing nothing is not an option.)
And the one path that aggressively claims otherwise 🔥 ( roughly, "Elon Musk hacks the simulation to invent propellant-free rockets") is completely fictional.
You can't just say "the future will have AI and be bad 💡 ( bad according to a myopic metric, with human population uber alles) , so we need to stop it".
Because:
- There are other metrics that matter
- All of the future paths have issues