Tech

Most approaches to AI alignment incur upfront costs to the creator (an “alignment tax”). In expectation, startups will skimp on alignment in order to avoid the tax. “Alignment windfalls” are strategies which tend towards long-term public good and short-term benefits to a company. One suggestion for a high impact career is to discover, validate, and advocate for these alignment windfalls so as to accelerate their adoption.

With the rapid increase in capability of coding assistants like Copilot, should they be banned from technical interviews?

It's useful to be able to preview the changes made in a PR before merging it. This is fairly easy with Argo CD, but not very well documented: I'll recommend an approach so you can get this set up quickly.

Want to make Artificial Intelligence less likely to result in doom? Don't have a machine learning PhD? You can still help!

Ought are an applied research lab, studying how to build capable AI-powered systems without the existential risk. I'm super excited to be a part of the team!

At one level: yes. At another: maybe.

A machine learning approach called CycleGAN has been shown to be effective at transforming an image from one style to another. Perhaps turning a photo into a Monet-style painting, or changing a picture of a horse into that of a zebra. However, I wanted to try to use the approach on a more challenging transformation: turning audio data into an aesthetically-pleasing image to illustrate the latent structure of music.

In The Moral Landscape, Sam Harris describes a framework which allows for strictly rational conversations about morality. In this post we explore the possibility of applying this framework to serve as the basis of ethical artificial intelligence, and show that although The Moral Landscape might work well for humans, it's a poor choice for machines.

Existential risk is hard to weigh up versus other priorities we see around us in the world. When we look at the data, however, I think it's clear that this deserves more of our attention, and is generally under-funded when compared to shorter-term charitable causes.

I talked with Joel from the Modern CTO podcast about building and scaling an organisation, cultural habit formation, Catalan micro-villages, social media, among other things!

There are striking similarities between the training of a neural network and the way that we – as humans – form habits. How should that inform our thinking about the potential for this type of artificial intelligence, and what we should and shouldn't be asking of it?

My default position is that roadmaps are bunk.

I've found Atomic Habits to be a useful tool for my day-to-day goals – I was wondering if they could also be applied at an team-level?

… in which I weigh in on a sensitive debate from a uniquely privileged and blithely ignorant perspective.

Don't video player controls seem dumb? Desktop and web-based players alike allow the user to jump forwards or backwards by 5 or 10 seconds, or possibly speed up and slow down playback. These controls operate identically no matter what the user is watching – there's no connection between the content of the video, and the controls. Here's an idea for how to identify the salient points in a video that a user might want to skip to: the good news is that this could all be done automatically, and with very little overhead; I'll leave the bad news to the end!

None of the major cloud infrastructure providers are going to work well for the unique capabilities of Elixir and BEAM.

An attempt to justify spending hours in my workshop, breathing in harmful wooddust and sustaining several fairly serious injuries.

At Teespring, we use a take-home test as part of our technical interview process. We've had very little push-back on this from candidates, so the level of antipathy towards take-home tests in some discussions was an enormous surprise. Most people viewed these tests as a waste of time, rude, disrespectful, or a combination of all three. Here's why we think they're great.

Summary: yes-ish…

Introducing "full-stop" a little projects which helps separate the mechanism of managing and appyling your dotfiles from the content of the dotfiles itself.

A quick and dirty proof of concept for a virtual reality Teespring campaign page, created as a potential hackathon project.

I was recently asked: what's the worst project you ever worked on? I settled upon my first job out of university – taking "Not Invented Here" syndrome to a whole new level.