Maksim Karliuk
@karliuk.com
Learning to philosophise public policy at LSE. PhD and LLM (not large language model😀). Author of one and a half books📚
www.karliuk.com Café-philo: https://t.me/cafephilo
www.karliuk.com Café-philo: https://t.me/cafephilo
or young enough not knowing?
July 30, 2025 at 10:01 PM
or young enough not knowing?
The idea is that this document will help to ensure that AI models, such as LLMs, are safer and more transparent. One particular concern I have is that these systems will likely determine the future of human agency and autonomy as their use spreads into more aspects of daily life.
July 11, 2025 at 11:28 AM
The idea is that this document will help to ensure that AI models, such as LLMs, are safer and more transparent. One particular concern I have is that these systems will likely determine the future of human agency and autonomy as their use spreads into more aspects of daily life.
but this requires further investigation, as this reply shows: www.nature.com/articles/s41...
Reply to: ChatGPT decreases idea diversity in brainstorming - Nature Human Behaviour
Nature Human Behaviour - Reply to: ChatGPT decreases idea diversity in brainstorming
www.nature.com
May 16, 2025 at 7:44 PM
but this requires further investigation, as this reply shows: www.nature.com/articles/s41...
the download link on this page doesn't seem to work
April 18, 2025 at 10:29 AM
the download link on this page doesn't seem to work
true, that is mostly historical, i eventually focus on desirability
April 18, 2025 at 10:23 AM
true, that is mostly historical, i eventually focus on desirability
this thread looks pretty solid. the point is indeed that VFI shouldn’t reject weighing of values. re epistemic priority: it would still be default, but far-reaching detrimental consequences might justify balancing in favour of non-epistemic values
April 18, 2025 at 10:21 AM
this thread looks pretty solid. the point is indeed that VFI shouldn’t reject weighing of values. re epistemic priority: it would still be default, but far-reaching detrimental consequences might justify balancing in favour of non-epistemic values