This is a weekly round-up of things I’ve posted in the last week.
InkHaven requires that I post a blog post every day, which is a lot. Especially for people subscribed to my blog. Someone requested I spare their inbox, so I haven’t been sending out every post.
So now you get to catch up! You can even be selective if you prefer :)
The posts are:
About the posts:
Diary of a “Doomer” (part 1) is about my experience getting into the field of AI and AI Safety (I started graduate school in 2013). A lot has changed since then. What used to be a fringe topic has become really mainstream! I’m talking about deep learning, of course… But seriously, AI researchers really dropped the ball, and owe society a debt they can probably never repay for failing to consider the consequences of their actions.
Contra Leicht on AI Pauses takes apart Anton Leicht’s piece arguing we shouldn’t try to pause AI. I first encountered Leicht when he was arguing against having an “AI Safety” movement at all last fall. I don’t think either of these articles are very good — I find the reasoning sloppy.
Post-Scarcity is bullshit is mostly about how certain things are fundamentally scarse; like land, energy, and status. I got a bit snarky here about the discourse around the topic, and how vague, incoherent, and/or unimaginative people’s visions of the “post-scarcity” world typically are.
From Artificial Intelligence to an ecosystem of artificial life-forms. If the AI race doesn’t stop, the natural end-point is the creation of artificial beings that proliferate, diversify, and radically reshape the world. This is one of my quick and dirty attempts to explain a part of my world view that really deserves a 30-page essay.
Idea Economics is a rare non-AI-related post about how and why I think people devalue ideas: Not because they’re easy to come by, but because they’re hard to hold on to if you share them. But then I ruin it by talking about the CAIS Statement on AI Risk as an example (it was sorta my idea).
Stop AI is an attempt to get the basic case for why we need to stop AI down in writing. It ended up basically just covering the risks and not why other solutions aren’t good enough (stay tuned, that might be the next post).
Stop AI Now argues against kicking the can down the road. I think that’s intuitively a bad idea, but here I give three particular reasons.
Commentary:
I did this as a bit of an experiment. Besides the person complaining to me directly, I did notice a dip in subscribers at some point after about seven posts in a row at the start. A blogger friend of mine with more of a following says they often lose followers after a post. I guess that makes sense… people don’t like their inbox being clogged.
I did still send out two of these posts as email notifications. The first one was deliberate, the second was an accident. You can see that the ones I sent out did get a lot more views. I’ll be curious to see how much this post makes up the difference!
Thanks for reading The Real AI! Subscribe for free to receive new posts and support my work.
Discuss