The paper “When Will AI Exceed Human Performance? Evidence from AI Experts” analyses predictions about progress in AI from researchers who published at the 2015 NIPS and ICML conferences. Only a quick survey was used to gather the data.
A quick survey does not evoke thoughtful forecasts. It provokes fast thinking and sampling from intuition. Daniel Kahneman’s “Thinking fast and slow” provides ample examples of how fast thinking is skewed by bias.
The paper even admits that different framing of the same question elicited vastly different predictions: the predicted time to full automation doubled depending on whether it was framed…
In August, at the United Nations’ Convention on Conventional Weapons, 80 countries met to ban autonomous weapons. The US, Russia, Australia, Israel, and South Korea blocked the attempt. Last week, Russia reduced the length of the next convention in 2019 to just one week. This will make negotiations harder.
Unlike the push against nuclear and chemical weapons, these countries say they want to explore the potential benefits of lethal autonomous weapon systems first. Most global powers, like the US, India, Russia, and China, are actively developing them.
Autonomous weapons will change warfare more than any other technology. They can act…
As I am gaining some experience in acute prolonged insomnia, I couldn’t resist the irony of reading Fight Club in my waking morning hours.
It is a very good book with some hefty writing and a good punch. The movie does not stray very far from it. Rewatching it, one notices that many lines are taken 1:1 from the book.
However, the book is much more obviously political than the movie. It is truly anti-consumerist, and it exposes the insanity and fakeness of the world we accept and live in. And, rightly so, you could say after reading it.
After two months as a fellow at Newspeak, I’ve realized that, while AI and ML are very interesting, they are not real game changers. The real game changers are cryptocurrencies and blockchain! AI and automation are still far away from transforming our society and will only further concentrate wealth in the hands of the few, whereas blockchain is happening *now*.
Blockchain can bring about decentralization and real fairness…
OpenAI’s gym environment only supports running one RL environment at a time. If you want to run multiple environments, you either need to use multiple threads or multiple processes. Moreover, this only parallelizes well as long as you have sufficiently many cores available. However, if your batch size increases beyond that, it won’t become any faster. The overhead also uses up compute that could be used for training and inference.
Most current RL implementations actually work by sampling training data from multiple environments and training on batches. This is usually done by having multiple workers that run multiple environments…
Before WaveNet, speech generation was suffering in the uncanny valley of “good enough to be understood but not good enough to sound pleasant” for a long time. Everyone is familiar with the robotic voices of parametric speech generation of old. WaveNet has changed all this. First published in a research paper by DeepMind in 2016, it was launched in Google Assistant in September 2017.
When Google Assistant replies to you, it uses a voice generated by WaveNet. The generated speech sounds much more pleasant now, and it has become harder to distinguish it from a real human voice. …
The Dart REPL allows you to evaluate Dart expressions and statements in an interactive shell. It has been awhile since my first post about the Dart REPL (you don’t need to read it to enjoy this post), and lots of features are still missing. In particular, dynamic imports and support for top-level declarations would be very useful, so let’s look into how to support them.
Disclaimer: I do work for Google, but this post is about a personal project. I’m not on the Dart team or related. This article only contains my humble personal opinion.
Disclaimer: The following only represents my personal opinion and is in no way related to my employer etc. Also I don’t know much, so please let me know when I’m wrong :)
Because Medium doesn’t support LaTex, I’m using a gist, which is a bit awkward. You can also enjoy this post as a blog post with MathJax: http://blog.blackhc.net/2017/03/dlb-chapter2/ or as a PDF: https://github.com/BlackHC/dlb_chapter2/raw/master/content.pdf
I’ve started reading the Deep Learning book. You can find an online version at http://www.deeplearningbook.org/. I’m trying to take notes and really interact with the material. This post contains my thoughts for Chapter 1. The ones that still make sense to me, that is.
Disclaimer: The following all represents my personal opinion and is in no way related to my employer etc. Also I don’t know much, so please correct me when I’m wrong :)
The main lesson of thirty-five years of AI research is that the hard problems are easy and the easy problems are hard.
Marovec’s paradox could…
Disclaimer: I’m just learning about proof verification systems. Bear with me on this journey :)
Recently, I’ve stumbled over Metamath. Metamath is a proof verification system. It allows you to formalize mathematical proofs and checks that everything is sound and consistent. Proof verification systems are related to automated proof solvers that try to go one step further. They want to automate the holy grail of the mother of all sciences. There are many interesting systems like Isabelle and Coq. However, they are quite complex. …
DPhil student at AIMS in Oxford; former RE at DeepMind, former SWE at Google; fellow at Newspeak House.