cmpalmer52

That’s a bit harsh. 57 here and my forearms, elbows, and wrists are f…ouled up without an ergonomic split keyboard and some other accommodations. Bodies are different, and if this guy is a rock climber, that might have something to do with it. But flat, tight keyboards give me tendinitis in as little as a day. I can’t even use my laptop without an external keyboard, which kinda defeats the purpose, but it’s what work gave me and it’s more powerful than my current, out of date, PC.

My dad does this, but he’s also hard of hearing so he can’t understand what they’re pitching half the time. So mostly he just answers so that he can cuss them out. Unfortunately, he’s done that to friends, medical offices, and relatives who call from an unknown number. Landline only, of course. I tell him not to answer, but he does anyway.

It reminds me of the 501 Mechanical Movements type books, which are endlessly fascinating to me.

Interviewed with a defense contractor. I had three “ins” at the company and was basically interviewing to check them out because I was about 90% sure I’d get an offer.

I was about to burn out at my current job at the time. Too much stress, too many hours, too much crazy. At this interview, every one I talked to talked about how “exciting” and “cutting edge” everything they were doing was. “We barely finish a prototype before they’ve changed the design and we have to start over without testing the last one.” “It’s very fast paced, stuff is changing all the time.” “We work over some. Last week, I put in 72 hours, but that’s not typical.” “A dozen companies have tried to solve this problem, but we think we’ve finally figured it out.” And so on. And it was a weapon technology, which I had ethical qualms about.

I called them back and withdrew my application. The HR guy said he was preparing my offer. I told him I didn’t want to see it. My biggest fear was that it would be enough money to tempt me to take the job and I would have been miserable.

Six months later, half of everyone there was laid off. They never got the big problem solved. They lost several contracts. Dodged that bullet (pun intended).

A universe is a computer that simulates a universe.

That may be the best description of the show, ever. My fondest memories of it were how detailed the online analyses of each episode were as they aired. All of the speculation, references to previous clues - everything was dissected. And JMS was right in the only fray most of the time, teasing and clarifying and explaining, but never spoiling what was coming up. It was the golden age of Internet synergy with a show, not just trolling and shitposting.

What we need are LLMs that can use the tools for mathematics (and other disciplines) that already exist and to write and execute new programs to hopefully solve novel problems, or at least to be able to glue the existing tools together with shells.

There’s a story by Jorge Luis Borges called “Funes the Memorious” about a man who remembers everything, but can’t generalize. There’s a line about him not knowing if a dog on the square glimpsed at noon from the side is the same dog as the one seen from the back at 12:01 or something like that. Swirls of smoke from a cigarette are memorized forever. He mostly sits in a dark room.

Campbell’s Cream of Chicken and Mushroom soup (that’s a type, although you can double the recipe with one can of Cream of Chicken and one of Cream of Mushroom) Can of water Add to saucepan Stir and bring to a boil Fill can with Minute Rice Stir instant into soup, remove from heat, cover Wait 8-10 minutes

This was a go-to college meal. It’s great on a cold day, very filling.

One day, I made mushroom risotto. I used the right kind of rice. I ladled the made-from-scratch stock over the rice one ladle at a time, stirring slowly. I added the mushrooms and kept stocking and stirring. Finished with some fresh Parmesan. Took about 40 minutes of near constant attention and it was absolutely delicious.

Pareto Analysis:

The Campbell’s soup and Minute Rice is about 80% the same with minimal effort.

I have a physical HP-15C in my desk and I use the HP-15C emulator on my phone. As you can probably tell, I used that calc in college engineering classes until mine got stolen. I replaced it with an HP-28S that I used for quite a while. I bought the physical 15C a few years ago - 30 years after my original was stolen.

Wow, more JetBrains complaints than I expected. I’ve been using Rider for C# development for a while now (moved from Visual Studio - I’ve never been a VSCode fan) and, for the most part, it’s been worth it and my code has been the better for it. The analysis and linting is fantastic.

So far, the AI assistant in the EAP is one of the better AI integrations I’ve used in an IDE. I’ve used it for refactoring, simplification, generating XML doc comments, finding potential problems, but not much new code generation. I’ve also got Copilot, which mainly serves as a smarter autocomplete.

You do have to be careful - I’ve had it miss parameters on a refactor, but usually telling it “you missed parameter xyz” and it will correct it. But, hopefully, that’s just growing pains and it will improve.

The average person has no idea what 90% of the headlines on HN are even referring to, giving it a degree of self-moderation.

I did this for a while with my Quest 2. Floating in space on an asteroid while my main screen is bigger (virtually) than an IMAX screen and two other monitors float on either side plus tool windows I could grab and move around was awesome. Just the headset was heavy and sweaty after a while and there was no AR pass thru, so I couldn’t see my keyboard and mouse or find my coffee cup. This looks like a step forward in solving all those problems.

I’m probably in the smaller majority of people who are very excited about it, but still expect it to flop (or at least flounder for a while).

I used a Quest 2 as virtual monitors for a week or so of real work and it was uncomfortable and sweaty and I was isolated and had to keyboard and mouse by touch, but the viewing experience was amazing. And I do use my Quest 2 frequently for games and “workouts” and entertainment. So I’m solidly in the market for this.

But the main thing is that I want to try it, mostly for the AR. I’ll reserve any fanboy comments until I’ve tried it.

Someday, someone is going to get AR/VR right. This isn’t it, but it seems to be moving in the right direction.

I don’t have high hopes for free space holographic displays, a la Star Wars or Minority Report, but augmented reality fills the same niche (with many of the same drawbacks and benefits).

It sounds like a great idea to me. I’ve been trying to carve out time to do more electronic tinkering and learning and that sounds like a neat set of projects.

Does it work on a Raspberry Pi? I’ve got a broken antique table radio that I’ve been thinking of hacking into an Internet radio. I figured I’d put an encoder on the tuning knob and have it scroll through a set of presets.

Yes, I know it’s not modeling our neurons and connectivity directly, but it’s closer than traditional software or “expert systems” approach. And, lo and behold, we’re seeing emergent behavior that seems to resemble some of quirks of our meat network.

Even the mistakes it made seem like the kind of mistakes people make, not the kind of mistakes software usually makes.

This has been the plaguing my thoughts recently as well. When I analyze (or attempt to analyze) what’s going on in the black box of gray matter in my head, I’m increasingly aware that it’s nothing magical. I’ve stated to others that the real shakeup the latest AI successes may cause won’t be that we’ve created an artificial mind, but that our own minds may be nothing more than a collection of neural subsystems. Subsystems process our vision and pass it on. They process our hearing and pass it on. They process our senses of proprioception and pass it on. Then our cortex processes all these inputs and produces outputs, which may be speech, physical reactions, or just thoughts that loop back in as another input to the cortex. Maybe our consciousness is just a supervisory neural network that monitors what the cortex is doing and feeds in adjustments. But instead of prompt/response, it’s always running, feeding back on itself, trying to predict what’s going to happen next and adjusting its models based on what we observe. It explains a little too much and I start wondering if there really is a “me” that can be described as a conscious personality and I get a little creeped out. One consolation is that we’ve always equated some new technologies with “maybe that’s the way the mind works.” The mind is like a book. The mind is like a machine. The mind is like a calculating device. The mind is like a computer. And now we’ve reached: The mind is like a super LLM with assorted subsystems. The difference is, this time, we may be right since, for really the first time, we’ve started simulating it from the bottom up.

Maybe our personality and being is just a result of our training data with a few genetic quirks thrown in. And what am I really doing when I “decide” or “create” something? Is it that much different from a LLM’s generation of language? Or an AI image generator?

One of the first things I tried (which I thought would be harder) was “Write a Python program that accepts a YouTube url and a start and stop time in the form MM:SS and a file name and it extracts the video between those times and converts it to an animated gif. Also provide a list of packages to be installed.”

The resulting code worked. Took an interaction or two to add usage info and tweak things, but it’s a neat little utility. Due to the libraries used, it was also simpler than I expected and would have been easy to write if I knew about those libs, so I also learned something.

It’s just one or two steps away from just saying “Go to this YouTube URL and extract the video between 3:20 and 3:27 into an animated GIF named ‘CatAndRaccoon.gif’” and having it write, debug, and execute the code.

This one was using Bing AI chat:

I had a bunch of PDF files with coded file names, like ABC-123-abc-999-001.pdf, where each section of the file name was meaningful. Inside the PDF were several form fields. I needed to insert records in a database for each of the 97 files. Easy, but tedious.

I prompted it with a description of the file name breakdown, where the text to grab was in the PDF, and then asked for a Python program to find the files (in subdirectories), extract the PDF text, and write a text file with the SQL Insert statements for all the files.

It took two or three minor iterations, but less time than it would have taken me to write from scratch because my Python is rusty. Regular expressions, PDF processing libs, file system traversal, and SQL generation, and it all compiled and worked from the start (the iterations were to tweak a few things I didn’t specify).

This is the kind of thing that I think is perfect for these tools. With the new tools that let the LLM compile and execute code, it will be cool (and potentially dangerous).

I’ve been very happy with most questions and coding queries. The worst results I’ve got were ones related to external libraries. We use Telerik UI for Xamarin components and Telerik has similar libraries for ASP.net, WinUI, JavaScript, etc. and the controls often have the same in different libraries and different methods and properties. Try as I might to steer it, it has yet to produce any code using these components that doesn’t mix methods. And as others have said, it will argue with you about those features or it will say “Sorry, that method is for the WinUI version. Here’s a corrected version:” and then make the same mistake again.

It is poorly written. I assume it’s because Fantastic and Incredible are “unobstructed, free fall pits” and Smokey I is partially obstructed or not a clear vertical drop in some way.

Yeah, that seemed a little hand-wavy in its description. Thanks for reviewing, it’s been very interesting.

Surprisingly, I understood some of that (not a strong math background), but I asked Bing AI to explain it at the level a smart high schooler would get it. How did it do?:

Let me try to restate this in a simpler way. A polynomial is a mathematical expression that involves adding and multiplying powers of a variable, such as x. For example, x^2 + 3x - 5 is a polynomial of degree 2. A coefficient is the number that multiplies a power of x, such as 3 in 3x. A coefficient vector is a list of all the coefficients of a polynomial in order from highest to lowest power of x, such as [1, 3, -5] for x^2 + 3x - 5.

Sometimes, we want to know what happens when we plug in different values of x into a polynomial. For example, what is the value of x^2 + 3x - 5 when x = 2? This is called evaluating the polynomial at a point. We can also evaluate the polynomial at many points at once, such as x = 1, 2, 3, 4. This is called evaluating the polynomial at a set of points.

The Fourier transform is a technique that helps us do this faster and easier. It takes a coefficient vector of a polynomial and transforms it into another vector that contains the values of the polynomial at a special set of points. These points are called powers of a primitive Nth root of unity. This sounds complicated, but it just means that they are numbers that when multiplied by themselves N times give 1. For example, when N = 4, one such number is i (the imaginary unit), because i^4 = 1. The powers of i are i^0 = 1, i^1 = i, i^2 = -1, i^3 = -i. These are the four points that we can evaluate any polynomial of degree 3 or less using the Fourier transform.

The inverse Fourier transform does the opposite. It takes a vector that contains the values of a polynomial at some points and transforms it back into the coefficient vector of the polynomial. This is useful when we want to find the polynomial that passes through some given points. This is called interpolating the polynomial from the points.

The Fourier transform and the inverse Fourier transform are very similar, except that they use different orders of the coefficients and the points, and they have to divide or multiply by some number (the scaling factor) to make sure everything works out.

The nice thing about finite fields is that they always have some numbers that can be used as powers of a primitive Nth root of unity for any N. A finite field is a set of numbers that has a fixed size and follows certain rules for adding and multiplying. For example, GF(7) is a finite field with 7 elements: 0, 1, 2, 3, 4, 5, 6. In GF(7), we add and multiply by taking the remainder after dividing by 7. For example, 4 + 5 = 9 mod 7 = 2 in GF(7). In GF(7), one primitive Nth root of unity for any N is 3, because 3^N mod 7 = 1 for any N that is divisible by 6. The powers of 3 are 3^0 = 1, 3^1 = 3, 3^2 = 9 mod 7 = 2, ..., and so on.

This means that we can use the Fourier transform and the inverse Fourier transform to evaluate and interpolate any polynomial in any finite field without worrying about finding the right points to use.

I hope this helps you understand better what the original text was saying. If you have any questions or want more examples, feel free to ask me.

At this point, with the boom in AI, I would settle for very simple AR glasses. Not trying to make 3D objects meld with the the real world view, but just superimposed graphics, kinda like Google Glass, but better. Imagine if it could automatically subtitle foreign languages, or translate Japanese street signs and advertising in real-time. Join that with a conversational AI so you could point your eyes at something and just say “What is that?” and then get whatever level of detail response you want, from a subtitle, to a voice description, to a full dive into the historical details. This is the market space for a fashionable device, but it needs to be a real view of reality, not video pass-through. Imagine renting one (if they’re expensive) when traveling to a foreign country, or at a museum. The cool think about AI integration is that you wouldn’t have to create the content like an audio-guide does. In a museum, you could learn about the artist, the time period, critical analysis, the school of art, etc. dynamically and organically. As a tourist, you could look at an attraction across the street, read (or hear) reviews, check hours and prices, book tickets, etc. Would also be great for directions.

Agreement from me. I have had no complaints about the Roku TV experience on my TCL.

I write software for natural gas utility management and maintenance. Between this and coding AIs, I’m glad I’m approaching retirement age.

Of course, new government regulations (not outright bans) are a lot of what drives the need for our software.

Same. I’ve been using it quite a bit for work, but I’ve also been just playing/experimenting with it and many of my questions have come from knowing a tiny bit of Japanese and watching anime. It’s great for explaining differences between words or phrases that translate similarly.

A friend attacked a patch of undergrowth with a string trimmer (“weed eater”). She was wearing shorts and the pulped fragments of poison ivy splattered all over her legs and arms. Imagine putting poison ivy in a blender, then pouring it on yourself.

I use a 60% keyboard that doesn’t even have dedicated arrow keys. My caps-lock is configured as the Fn key and, when it’s held down, the JKLI letters are the arrow keys. This (and other macros) are actually really convenient when you get used to them. Unfortunately, it’s hard for me to type on “normal” keyboards because when I end up tiYping thinks like that. My keyboard is a Mistel MD600 split.

My favorite unusual accessory is a little 9 key mechanical auxiliary keyboard. I programmed it to be my debug keyboard with all of key combos for my IDEs debugger - run, stop, pause, next, step over/into, set breakpoints, etc. it saves a lot of time (or seems like it does).