After getting a bit lost in the question of whether we should in principle be able to have a clue about future ramifications of our actions, I would like to turn back to questions of how well we're actually doing so far. It looks like, for long-term projections, we are indeed pretty clueless. This document, … Continue reading How good are we at predicting the future?
"Another writing challenge!?" you ask. Fair enough. Maybe I'll drop the word "challenge", to make it seem less grandiose. But it definitely seems to be the case that I deal better with work/research/thinking/writing sprints than with trying to produce a constant trickle of output. Even when I remembered that "write it down" was a good … Continue reading Zooming in: Writing Sequence Number Two
Are humans machines; frail bodies to be optimized and overcome? And what is this haunting sense of there being more to it? An exploration through the cocktail parties of my mind.
This morning, I read the 2014 FHI policy brief "Unprecedented Technological Risks". Unusually critically-minded today, I found it easy to notice some assumptions that were at least non-obvious to me, which in turn might point either to them lacking justification, or me not having looked at them well enough. Weird assumptions I don't have much … Continue reading Unclear assumptions in AI risk
What are the primary risks to society of failure for AI systems, and how can these risks be monitored and addressed at scale?DeepMind Ethics & Society I found myself cringing away from the question, while at the same time dismissing it - a sure sign that I should look at it for at least half … Continue reading AI risk: bullets, bullet-points, and poems
I am sorry. This is an awfully broad question, but it's already late (which I notice by counting the number of spelling mistakes I make in a sentence - 4 in this one so far, ah, now 5). I had lots of thinking time and reading time and discussion time today. My head is one … Continue reading What do we need to forecast AI development?
Trying to get a handle on how neural networks and deep learning work. Also how to fool a neural net, and speculations about the future of AI.
Would machines that can "do language" have general intelligence?
I seem unable to answer the former question without reference to the latter. I've been a bit confused about what it means to be an agent. I also know that other people are spending actual time doing actual philosophy on the agent vs. tools question, but that shall not detract me from starting some preliminary … Continue reading How likely are AIs to be agents? What does it mean to be an agent?