I recall an earlier exchange, posted to HN, between Wolfram and Knuth on the GPT-4 model [1].
Knuth was dismissive in that exchange, concluding "I myself shall certainly continue to leave such research to others, and to devote my time to developing concepts that are authentic and trustworthy. And I hope you do the same."
I've noticed with the latest models, especially Opus 4.6, some of the resistance to these LLMs is relenting. Kudos for people being willing to change their opinion and update when new evidence comes to light.
DF gets all the news (rightfully so, it's an epic game that I've dumped a ton of hours into) but if you haven't already, consider checking out Songs of Syx. It's like DF but multiplied by 100. You can have tens of thousands of citizens, doing most of the things they do in Dwarf Fortress, and a lot more, including waging huge wars against the neighbors. The limits of DF kinda made me sad, actually, that you are limited to so few Dwarves (and don't say it's because you want to know the story of all of them, because after 30 or so you lose track of who is who anyways, so might as well up the limit from 100 to 50K, or more? ;) Songs of Syx has also routinely been getting massive updates since 2020 and I have a feeling the code is a bit cleaner so the solo dev can add features faster (unlike DF's code base which is, according to one of the new devs a nightmare to work with). It's a game that is never talked about but deserves a whole lot more love from gamers.
I don't mean to cast shade on DF, I really do love it, and am happy for its existence, I just think that DF fans should also look into Songs of Syx.
The defining difference for me are the generated stories in DF, which often are a lot of random trash but still give a feeling of a deeper meaning.
It’s probably the easiest way of setting up a Mac VM if you have unraid. I know there are similar options for qemu and kvm based hypervisors. If you have an amd gpu you should be able to pass it through.
I use "about the same", thanks to a friend. I love the reactions (from Americans, where everyone is expected is to say "Great" or "Good" or something similarly positive).
My wife was right too. My kids ended up being unable to manage their device use at all, they developed seriously bad habits, lied and deceived extensively to gain access to devices, and repeatedly sacrificed relationships and trust for more screen time. There were years there where I thought surely they'd click with it and develop better habits, make better choices (with our guidance), and so on. Abstinence could be worse, right? Some exposure would be helpful and lead to useful conversations and so on.
The Internet, Internet access, and apps have changed since I was a kid. Despite their time on digital devices along with my efforts to teach them, my kids have no idea how computers work or how to use them very effectively. The skills they have developed to gain access to them were largely social engineering and lying. They exclusively waste time and brain cells when they're on screens.
One of my kids essentially can't have access to devices because he'll burn hours into the night playing really, really stupid games and watching porn. This is ALL he wants to do on phones or computers. Sometimes he will window shop.
You might think this is largely due to my failure to have insight into what my kids are doing and limiting access correctly, but that isn't the case. At first we were somewhat lenient and figured if they accessed things they shouldn't, we'd see it and have conversations. That was very early on. The conversations did nothing. I began putting severe restrictions on devices quite quickly because problems became evident quickly. I was a bit naive about it at first, my wife was not. We clashed a bit, but then device theft and social engineering started and I quickly aligned with her. Since then, many years ago, very little access has been on account of us not protecting devices properly. He is extremely good at gaining access when he's not supposed to, and extremely good at hiding it. It's like having an addict in the house.
He has no future in computers. He doesn't care about computers at all. He is incredibly compulsive, self-harming, and freely harms his relationships to get what he wants. This has been going on for about 5 years; he's 16 now, and I'm pretty scared for when he's out on his own and doesn't have anyone to protect him from himself. I think there will be some brutal lessons. Lost jobs, lost relationships, lost confidence and self esteem. I'm not looking forward to it.
I have no idea why I turned sneaking onto computers into a career rather than rotted away like they do. I wanted to learn to program. I was curious. My kids want to play NBA 2k and watch porn. That's about it.
The Netherlands in 2025 is a decadent country were everyone can do whatever the hell they want.
But a gay man growing up in the 1950s in a rural village was plenty oppressed.
It's actually quite fascinating how in the 1960s/70s we had a Cultural Revolution of our own that ended a thousand years of religious oppression! And we didn't even have a Mao.
But never forget we are always one bad week away from sliding backwards.
Since I might have peoples' attention for a moment (Can no longer edit my original post). Might I suggest the King William's college quiz this Christmas?
We have successfully replaced thousands of complicated deep net time series based anomaly detectors at a FANG with statistical (nonparametric, semiparametric) process control ones.
They use 3 to 4 orders lower number of trained parameters and have just enough complexity that a team of 3 or four can handle several thousands of such streams.
The amount of baby sitting that deep net models needed was astronomical, debugging and understanding what has happened quite opaque.
For small teams, with limited resources I would still heavily recommend stats based models for time series anomaly detection.
May not be your best career move right now for political reasons. Those making massive bets do not like to confront that some of their bets might not have been well placed. They may try to make it difficult for contrary evidence to become too visible.
> It is well known the drunken sailor whos taggers to the left or right n independent random steps will, on the average, end up about sqrt(n) steps from the origin. But if there is a pretty girl in one direction, then his steps will tend to go in that direction and he will go a distance proportional to n. In a lifetime of many, many independent choices, small and large, a career with a vision will get you a distance proportional to n, while no vision will get you only the distance sqrt(n). In a sense, the main difference between those who go far and those who do not is some people have a vision and others do not and therefore can only react to the current events as they happen.
Just a tiny bit of bias towards a direction will get you very far very fast.
I once modeled+visualised this with a bit of javascript[1] and it's quite surprising to see the huge difference from even a tiny multiplication factor on each random/probabilistic decision.
Currently about 2,000 people play every day and I’ve released 59 puzzles!
One feature I’m excited about is crowdsourcing puzzles. Today’s puzzle is a “community puzzle” made entirely from clues that players submitted! I plan to do this every week or two.
I wrote about launching and the first month of puzzles if you want to learn more!
An PWA primarily for my wife and my daughter. They can order their hot chocolate and their coffee as if they were going to grab something at a fancy café downtown, but instead it's at home and I'm the barista. It is quite nice to have for when my wife comes back from work and want something specific, or when we are waiting for the visit of a few friend, they can order exactly the available beverages and everything is ready when they're here.
It was also a good playground for me to implement Web Push notifications (to never miss new orders).
It's a basic Nuxt 3 app with Appwrite as the backend with rough edges, but much enough for our household use !
I did something similar many years ago. I fed about half a million words (two decades of mostly fantasy and science fiction writing) into a Markov model that could generate text using a “gram slider” ranging from 2-grams to 5-grams.
I used it as a kind of “dream well” whenever I wanted to draw some muse from the same deep spring. It felt like a spiritual successor to what I used to do as a kid: flipping to a random page in an old 1950s Funk & Wagnalls dictionary and using whatever I found there as a writing seed.
This keeps coming up and we keep having the same debates about what Age Verification isn't.
For the folks in the back row:
Age Verification isn't about Kids or Censorship, It's about Surveillance
Age Verification isn't about Kids or Censorship, It's about Surveillance
Age Verification isn't about Kids or Censorship, It's about Surveillance
Without even reaching for my tinfoil hat, the strategy at work here is clear [0 1 2]. If we have to know that you're not a minor, then we also have to know who you are so we can make any techniques to obfuscate that illegal. By turning this from "keep an eye on your kids" to "prove you're not a kid" they've created the conditions to make privacy itself illegal.
VPNs are next. Then PGP. Then anything else that makes it hard for them to know who you are, what you say, and who you say it to.
Please, please don't fall into the trap and start discussing whether or not this is going to be effective to protect kids. It isn't, and that isn't the point.
Even though the article is mostly talking about visualizations, but I thought I'd share that I did at one point build a dance choreography software that renders the UI entirely SVG. I was surprised as to how well that worked.
If you're curious, it's called StageKeep, and you can find it here. https://stagekeep.com/
The original project used React Three Fiber, but refactored it to SVG for reasons I don't quite remember. I was inspired by signed distance functions, and the fact that one function could have such an outsized visual effect. Although the software doesn't use SDFs, but I like the idea of atomic functions that accepts some input, and outputs SVG.
I feel like I’ve figured out a good workflow with AI coding tools now. I use it in “Planning mode” to describe the feature or whatever I am working on and break it down into phases. I iterate on the planning doc until it matches what I want to build.
Then, I ask it to execute each phase from the doc one at a time. I review all the code it writes or sometimes just write it myself. When it is done it updates the plan with what was accomplished and what needs to be done next.
This has worked for me because:
- it forces the planning part to happen before coding. A lot of Claude’s “wtf” moments can be caught in this phase before it write a ton of gobbledygook code that I then have to clean up
- the code is written in small chunks, usually one or two functions at a time. It’s small enough that I can review all the code and understand before I click accept. There’s no blindly accepting junk code.
- the only context is the planning doc. Claude captures everything it needs there, and it’s able to pick right up from a new chat and keep working.
- it helps my distraction-prone brain make plans and keep track of what I was doing. Even without Claude writing any code, this alone is a huge productivity boost for me. It’s like have a magic notebook that keeps track of where I was in my projects so I can pick them up again easily.
People want the cookie, but they also want to be healthy. They want to never be bored, but they also want to have developed deep focus. They want instant answers, but they also want to feel competent and capable. Tech optimizes for revealed preference in the moment. Click-through rates, engagement metrics, conversion funnels: these measure immediate choices. But they don't measure regret, or what people wish they had become, or whether they feel their life is meaningful.
Nobody woke up in 2005 thinking "I wish I could outsource my spatial navigation to a device." They just wanted to not be lost. But now a generation has grown up without developing spatial awareness.
Make the model scale to be 10000000 (10 million). The sun is a chunky 139 meters in diameter. Earth is 15 km (9 miles) away. Pluto is 587 km (365 miles) away. The speed of light is 107 kph (67 mph).
Alpha Centauri is 4.1 million km (2.5 million miles) away... that is 10 times the earth moon distance.
Another comparison... Voyager 1 is moving at 30 light minutes per year. (Andromeda galaxy is approaching the Milky Way at 3.2 light hours per year)
Nice article, though the diagram showing [OpenGL] [WebGL] [WebGPU] being built on Vulkan and then from Vulkan to D3D12 and Metal is wrong. WebGL and WebGPU go directly to D3D and Metal, they do not go through Vulkan first
Also, Vulkan is labeled as Open Source. It is not open source.
The are other mistakes in that area as well. It claims WebGPU is limited to Browsers. It is, not. WebGPU is available as both a C++ (Dawn) and a Rust (WGPU) library. Both run on Windows, MacOS, Linux, iOS, and Android. It is arguably the most cross platform library. Tons of native projects using both libraries.
Vulkan is also not really cross-platform any more than DirectX . DirectX runs on 2 platforms (listed in the aritcle). Vulkan runs on 2+ platforms, Android, and Linux. It runs on Windows but not on all windows. For example, in a business context using Remote Desktop, Vulkan is rarely available. It is not part of Windows, and is not installed by default. Graphics card companies (NVidia, AMD) include it. Windows itself does not. Vulkan also does not run on MacOS nor iOS
Knuth was dismissive in that exchange, concluding "I myself shall certainly continue to leave such research to others, and to devote my time to developing concepts that are authentic and trustworthy. And I hope you do the same."
I've noticed with the latest models, especially Opus 4.6, some of the resistance to these LLMs is relenting. Kudos for people being willing to change their opinion and update when new evidence comes to light.
1. https://cs.stanford.edu/~knuth/chatGPT20.txt