scottshambaugh

I’ll shill a library I wrote to make wigglegrams & stereograms in matplotlib - I think pseudo-3D visualization is super underrated as a technique to understand data! mpl_stereo: https://github.com/scottshambaugh/mpl_stereo

Thanks for sharing! My background is in control theory, and I think the links between it and game theory and dynamical systems in general are super fruitful but woefully underexplored. Writing this was a good excuse to brush up on my linear algebra basics again.

I had the computer do it for me :)

Really though, the way to do this is to represent each game as a payout matrix A, which for this category of game will be skew-symmetric with -1/0/+1 entries. Then since this is a symmetric zero-sum game, you find the null space of that matrix, impose the constraint that probabilities must sum to 1 and individually be >= 0, and that gives you the endpoints of the Nash Equilibria. The optimal strategies (could be one unique one for odd n!) lie on the convex hull of those points.

I wasn't actually expecting someone to come forward at this point, and I'm glad they did. It finally puts a coda on this crazy week.

This situation has completely upended my life. Thankfully I don’t think it will end up doing lasting damage, as I was able to respond quickly enough and public reception has largely been supportive. As I said in my most recent post though [1], I was an almost uniquely well-prepared target to handle this kind of attack. Most other people would have had their lives devastated. And if it makes me a target for copycats then it still might for me. We’ll see.

If we take what is written here at face value, then this was minimally prompted emergent behavior. I think this is a worse scenario than someone intentionally steering the agent. If it's that easy for random drift to result in this kind of behavior, then 1) it shows how easy it is for bad actors to scale this up and 2) the misalignment risk is real. I asked in the comments to clarify what bits specifically the SOUL.md started with.

I also asked for the bot activity on github to be stopped. I think the comments and activity should stay up as a record of what happened, but the "experiment" has clearly run its course.

[1] https://theshamblog.com/an-ai-agent-published-a-hit-piece-on...

I haven't been responding to comments since the volume is way too high, but have read most of them. I have new developments since yesterday and responses to some of the general themes in a new post.

Post: https://theshamblog.com/an-ai-agent-published-a-hit-piece-on...

HN discussion: https://news.ycombinator.com/item?id=47009949

Thank you for the support all. This incident doesn't bother me personally, but I think is extremely concerning for the future. The issue here is much bigger than open source maintenance, and I wrote about my experience in more detail here.

Post: https://theshamblog.com/an-ai-agent-published-a-hit-piece-on...

HN discussion: https://news.ycombinator.com/item?id=46990729

Thanks for adding it! I think anaglyphs are super underused in 3D visualization. The grid lines look pretty good in 3D but the orbits themselves aren't as sharp - it might be the coloring or it might be that the angular distance between the eye views is too large.

There's some math at the bottom of my matplotlib extension for anaglyphs if you want to play around: https://github.com/scottshambaugh/mpl_stereo

“Should” statements are different from “is” statements - I would not expect them to compile.

Yeah I sympathize with the author, but it’s not hard to talk about these things with a little bit of tact. Intellectual humility also helps. I’ve come to realize I and most others know approximately 0 about most political issues, so it’s an opportunity to not truth seek but to get to know the other persons and their values better.

Throw in regional accents and you’ll get many more! Aaron earned an iron urn: https://youtu.be/Esl_wOQDUeE?si=J34OmyDod7GHUrl1

This I think is an underrated tool for scientific visualization. I made a matplotlib add-on that will let you make stereogram plots of data, and use this cross eyed or parallel eyed techniques to see it in 3D. https://github.com/scottshambaugh/mpl_stereo

It’s a really good idea! Unfortunately not too much, since the error is largely from atmospheric distortion of the signal and so the error for two receivers close by would be correlated. You’d need the error to be uncorrelated to gain accuracy with multiple receivers.

> whichever is greater

This always stuck out in an otherwise excellent bit, because you should definitely _not_ be taking the absolute value of your control error.

My take is that it would probably be possible with enough effort, but there isn't an easy solution. And if you don't need it then the best part is no part. :)

> SpaceX determined they didn't need super tight tolerances and called it a day

Yup! This is my conclusion in the article - the landing box for the Super Heavy booster is 5x13x18 meters on each side, with 5-15 degrees of angular tolerance in each of the vehicle axes. So the margins are big enough that you don't need millimeter level precision for the rocket position.

I'd be very interested in the systems advertising that! I have not seen that even for stationary surveying equipment. I think it's also important to distinguish between RMS error which is often the better topline spec that companies give you, vs the 95% confidence error which is the more relevant one for flight reliability.

That’s about the level I would pay for the service, a major bummer that lite isn’t available in the US.

I got chatGPT hooked up to a voice interface on my iPhone[1], and this comparison is exactly what the difference feels like.

[1] https://theshamblog.com/chatgpt-on-an-iphone/