TaylorAlexander

I was very surprised you didn’t make some comment about it! It had to be going through every reader’s mind. Happy to provide that comedic release for those readers.

I hate it when the TV signal is upside down because it is coming from Australia!

It’s very hard to change your standard of living, and the standard of living in California is very different. We live off of tens of gallons of water a day or more. When you have to cut back, it’s difficult for your human animal to accept it. Population wide, it’s nearly impossible.

Well, bicycles have more demand and volume (presumably), so that would play in to it. But yes, bicycles are a good point of reference.

I don’t know if that’s what people are saying, tho maybe part of convincing the public to fund such efforts involves strongly suggesting that could be the case. But what else can we do but poke them harder? It’s been working so far.

I have not followed them but I got the impression that AI simply doesn’t do what they said theirs could do.

> The policy is to move all the dirty work to someone else's back yard.

Whose policy? Like, which politicians and what exactly is their policy?

Because that sounds more like the kind of thing people say they are trying to do, while hiding the fact that they are doing something different.

I suspect that the extent to which they have done so, and its impacts, vary significantly from how things have gone in the US.

I actually think the extreme density and breadth of manufacturing in China is going to continue to outpace US manufacturing. They have multiple enormous manufacturing hubs connected by high speed rail lines over a wide geographic area. The US has no equivalent to the likes of Shenzhen and Guangzhou, where you can finish a PCB design in the morning and have the prototype in hand that evening. You can go to the Huaqiangbei Electronics Market and find exactly the right motor and controller for your specs, pick up specialized sensors, and build your next rev overnight.

I have lived in the Bay Area my entire adult life. We used to have Halted/HSC, we used to have Weird Stuff. We used to have Triangle Machinery Co in Santa Clara. Now everything is gone.

I think it’s great that we built a semiconductor manufacturing plant. That’s important for strategic manufacturing. But we’ve so thoroughly destroyed our manufacturing base, let the factories rot, and financialized property value that the “weird place with random electronics” can no longer even afford to do business. Starbucks makes more money, so in it goes.

US politicians love to shout about manufacturing. “Manufacturing jobs jobs economy growth.” But these people DO NOT understand how things get made. They have no serious industrial policy. They do not know the value of a high speed train connecting manufacturing centers. And even if they did, the entire apparatus of our government is set up to stop it.

Manufacturing workers need education. They need housing, transit, health care, maternity and sick leave. They need secure jobs and extra income that allows them time off to take classes to learn new skills.

I’m glad we passed the inflation reduction act, and the CHIPS act. We need that investment. But it’s going to take much more than that to “bring manufacturing back” and I’ve have seen time and time again that we do not have the vision or capability to move in the ways that would be required.

I hope manufacturing comes back. We desperately need it. But I’m quite frustrated that despite some marginal progress, the serious changes we need are not on the horizon nor seemingly beyond it.

Yeah that’s true, instead of small diameter wide rollers, you could have large diameter narrow rollers, like for example bicycle wheels or scooter wheels. Actually you could even use tracks for the rollers, which might be an interesting innovation. That would increase friction and reduce battery life, but there may be some advantages. Of course it would be very complex mechanically!

But something made with scooter wheels could be used to navigate around a city for example. That might be interesting.

I was a robotics software engineer on a project that built a mecanum wheeled robot. The big problem is that the rollers are small relative to the wheel. You know how bicycles handle rough terrain better than skateboards? The actual rollers on a mecanum wheel are pretty small compared to the diameter of the wheel they are mounted on. Our robot could handily carry a person on a flat floor (the engineers mounted a chair on a prototype and we got to sit on it and drive it with an Xbox controller - an actually extremely fun activity that could make a nice carnival ride). But if the robot encountered a pencil on the ground, the rollers couldn’t roll properly and things would get messed up.

I think I’ve seen suggestions of using the wheels for mars rovers, but a mecanum wheel that works in sand would need the rollers to be very large, so the actual wheel would be huge. That would suck up all your weight budget for your robot. Ackermann steering is lighter.

All that said, when you have a controlled environment with a flat floor, mecanum wheels are pretty rad!

Interesting. I assumed it was just sloppy journalism but it sounds like you’ve seen this confusion before. The name did not strike me as hard to remember.

I like that the article refers to the company as “Bonce Imaging” (image caption), “Bounce Imaging”, and “Bouncing Image”.

Bonce Imaging is definitely my favorite.

Right but there are people who use CAD for 3D printing projects around the house too. For them a few thousand a year is extreme.

The discussion around voter ID laws never actually revolves around expanding access to ID cards. Instead it is about restricting access to legal voters who don’t have ID, under the guise of avoiding voter fraud, a thing that is so rare as to be generally insignificant.

https://gspp.berkeley.edu/research-and-impact/policy-initiat...

Yah but if you want to click around on a neat lil web page this is great.

Nice, thank you!

I simply didn’t know the answer and the response “it would be much slower” is a perfectly acceptable reply. I disagree that it was ridiculous to ask. I was curious and I wanted to know the answer and now I do. What is obvious to you is not obvious to other people, and you will get nowhere in life insulting people who are asking questions in good faith.

Yeah. I can’t afford a freight train.

> It's kinda hard to believe that someone would stumble onto the landmine of AI performance comparison between Apple Silicon and Nvidia hardware.

I encourage you to update your beliefs about other people. I’m a very technical person, but I work in robotics closer to the hardware level - I design motor controllers and Linux motherboards and write firmware and platform level robotics stacks, but I’ve never done any work that required running inference in a professional capacity. I’ve played with machine learning, even collecting and hand labeling my own dataset and training a semantic segmentation network. But I’ve only ever had my little desktop with one Nvidia card to run it all. Back in the day, performance of CNNs was very important and I might have looked at benchmarks, but since the dawn of LLMs, my ability to run networks has been limited entirely by RAM constraints, not other factors like tokens per second. So when I heard that MacBooks have shared memory and can run large models with it, I started to notice that could be a (relatively) accessible way to run larger models. I can’t even remotely afford a $6k Mac any more than I could afford a $12k Nvidia cluster machine, so I never really got to the practical considerations of whether there would be any serious performance concerns. It has been idle thinking like “hmm I wonder how well that would work”.

So I asked the question. I said roughly “hey can someone explain why OP didn’t go with this cheaper solution”. The very simple answer is that it would be much slower and the performance per dollar would be 10x worse. Great! Question answered. All this rude incredulousness coming from people who cannot fathom that another person might not know the answer is really odd to me. I simply never even thought to check benchmarks because it was never a real consideration for me to buy a system.

Also the “#1 topic in the tech sector right now” funny in my circles people are talking about unions, AI compute exacerbating climate change, and AI being used to disenfranchise and make more precarious the tech working class. We all live in bubbles.

What? No I just don’t know the difference, sorry. I am interested in learning more about running 405b parameter models, which I believe you can do on a 192gb M series Mac.

The answer here is that the Nvidia system has much better performance. I’ve been focused on “can I even run the model” I didn’t think about the actual performance of the system.

I can't quite discern the subtext here but I will say that for many people international travel is not an appropriate solution for access to health care. People who are elderly, very sick, or without a passport, travel could be infeasible.

And beyond the logistical challenges, I think it should be fair to say axiomatically that international travel is not an adequate solution to healthcare access.

The other gas stations don't play pop, they play ads. The point is this business made a choice not to bombard people with ads, and I wish more places did that.

Those gas station ads are so offensive to the senses. I want so desperately for them to stop.

I used to live near Cupertino, and the Valero station on De Anza and Prospect always played the local classical radio station. Driving a car is pretty tiring, and to take a break and have loud ads projected at you only adds to the stress. To instead step out of your car to modestly amplified classical music is really much better. I always made a point to stop at that specific station for gas. We really need to be more mindful of the world we create for one another.

My friend has the endless captcha loop on ChatGPT too. Does anyone know how to fix it?

It says my iPhone 12 Pro Max doesn’t have WebGPU, but I enabled it in experimental features and another website[1] with WebGPU demos now works. Has anyone gotten this working on iPhone? Would be nice if the web app gave more info on what failed.

[1] https://webgpu.github.io/webgpu-samples/?sample=texturedCube

Well we will find out. I think the audience has value to advertisers, and I suspect that basic mechanic will continue in to the future.

> social media sites are already increasingly viewed by many people I know as having dubious value

I think we have all been saying this for 15 years but they keep getting more valuable.

Only if there are no hardware limits, which seems highly unlikely.

> If AGI is actually reached, capital as we know it basically becomes worthless. The entire structure of the modern economy and the society surrounding it collapses overnight.

Increasingly this just seems like fantasy to me. I suspect we will see big changes similar to the way computers changed the economy, but we will not see "capital as we know it become basically worthless" or "the modern economy and society around it collapse overnight". Property rights will still have value. Manufacturing facilities will still have value. Social media sites will still have value.

If this is a fantasy that will not happen, we really don't need to reason about the implications of it happening. Consider that in 1968 some people imagined that the world of 2001 would be like the film 2001: A Space Odyssey, when in reality the shuttle program was soon to wind down, with little to replace it for another 20 years.

I think LLMs are cool and that research should be focused on sample efficiency, building public copyright free datasets, and on recognizing and moving beyond the limits of LLMs. These systems are often very half-baked, and I am more interested in seeing research that moves to new ways of machine thinking than I am in dumping more and more GPUs in to this promising but incomplete technology.

As one example, Yann LeCun's vision of a system called JEPA [1] is interesting to me. It may not be the solution we need, but this type of thinking - taking what we have learned and exploring new architectures that may have even better real world performance - is what interests me.

[1] https://ai.meta.com/blog/v-jepa-yann-lecun-ai-model-video-jo...