Thinking Complete

"History is that certainty produced at the point where the imperfections of memory meet the inadequacies of documentation." Julian Barnes' novel The Sense of an Ending  (which won the 2011 Man Booker Prize) is short but fascinating; I recommend it. It also leaves a lot of questions unanswered or ambiguous; here's my take on what happened. Warning: many, many spoilers ahead. Adrian committed suicide because he'd gotten Sarah pregnant and she had decided to keep the child. The child, also called Adrian, had developmental issues because of Sarah's age, and eventually ended up in adult care. At the end of the novel, Tony feels responsible for Adrian's suicide because he remembers that he'd deliberately tried to set Adrian up with Sarah in order to sabotage Adrian's relationship with Veronica. He had done so by suggesting, in his letter to Adrian, that Adrian meet secretly with Sarah. Tony knew that Sarah would make a pass at Adrian becau

I think there are many questions whose answers would be useful for technical AGI safety research, but which will probably require expertise outside AI to answer. In this post I list 30 of them, divided into four categories. Feel free to get in touch if you’d like to discuss these questions and why I think they’re important in more detail. I personally think that making progress on the ones in the first category is particularly vital, and plausibly tractable for researchers from a wide range of academic backgrounds. Studying and understanding safety problems How strong are the economic or technological pressures towards building very general AI systems, as opposed to narrow ones? How plausible is the CAIS model of advanced AI capabilities arising from the combination of many narrow services? What are the most compelling arguments for and against discontinuous versus continuous takeoffs? In particular, how should we think about the analogy from human evolution, and the scalabi

Lately I’ve been thinking about how my thinking works, and how it can be improved. The simplest way to do so is probably to nudge myself towards paying more attention to various useful habits of mind. Here are the ones I've found most valuable (roughly in order): Tying together the act of saying a statement, and the act of evaluating whether I actually believe it. After making a novel claim, saying out loud to myself: “is this actually true?” and "how could I test this?" Being comfortable with pausing to reflect and thinking out loud. Trying to notice when my responses are too quick and reflexive, as a sign that I'm not thinking hard enough about the point I'm addressing. Asking for  specific examples , and using more of my own. Tabooing vague abstractions and moving away from discussing claims that are too general. Being charitable and collaborative, both towards new ideas and towards conversational partners. Trying to rephrase other people’s arguments and

I recently finished reading The Technology Trap , by Carl Frey. The book attempts to do two things: chronicle the role of technology in economic progress throughout history, and argue that automation in our own era parallels the first seven decades of the industrial revolution, during which the wealth from mechanisation failed to reach most citizens, leading to a populist backlash. I particularly enjoyed the first component, because until now I’ve read much less about the industrial revolution than I should have. That also means that I’m not qualified to evaluate the book’s accuracy. However, it had interesting discussions of: The technological prowess of the Romans, and why they were held back from industrialising both because of their slave-based economy, and also because of an implicit dismissal of the private economy. The development of some surprisingly important technologies during the Middle Ages, such as wind- and water-mills, better ways of accessing horsepower (via im

In his book The Fabric of Reality , David Deutsch makes the case for prioritising the explanatory  role of scientific theories over their predictive  role. I won't discuss this argument in detail, but a thought experiment given by Deutsch is fairly compelling. Suppose we had an “experiment oracle” that could predict the result of any experiment, but couldn’t tell us why it would turn out that way: “If we gave it the design of a spaceship, and the details of a proposed test flight, it could tell us how the spaceship would perform on such a flight. But it could not design the spaceship for us in the first place. And even if it predicted that the spaceship we had designed would explode on take-off, it could not tell us how to prevent such an explosion. That would still be for us to work out. And before we could work it out, before we could even begin to improve the design in any way, we should have to understand, among other things, how the spaceship was supposed to work. Only the

After about a decade of collecting and appreciating poems, I’ve decided to put together an anthology of my favourites, along with some commentary on what I like about each. My tastes lean towards short poems, and ones with an immediate poignancy or beauty to them - with a few notable exceptions which you’ll see below. I’m now 25, so it seemed appropriate to include 25 poems, in three sections. In general, poems are only ever about love or loss, and usually both, so the first two sections deal with those two themes. The third section consists of those rarest of creatures: non-romantic poems which are actually (mostly) happy. I hope you enjoy all of them as much as I have. Love poems: since feeling is first, by e e cummings No second Troy, by William Butler Yeats Marrysong, by Dennis Scott How do I love thee, by Elizabeth Barrett Browning A quoi bon dire, by Charlotte Mew La figlia che piange, by T. S. Eliot He Wishes for the Cloths of Heaven, by William Butler Yeats Jenny Ki