Sounds similar to backpacking through {India | Camino de Santiago}.
I think it's important for every young adult who becomes a well-rounded adult to have experienced a short term of deprivation so they have a frame-of-reference what others in less fortunate situations experience. The problem today is that there are too many mean, spoiled individuals with way too much power lacking theory of mind, a sense of community, and basic human compassion.
That is the key. The RPi works for idling, but anything else gets throttled pretty bad. I used to self host on the RPi, but it was just not enough[1]. Laptops/mini-PCs will have a much better burstable-to-idle power ratio (6/3W vs 35/8W).
"They're super expensive pattern matchers that break as soon as we step outside their training distribution" - I find it really weird that things like these are seen as some groundbreaking endgame discovery about LLMs
LLMs have a real issues with polarisation. It's probably smart people saying all this stuff about knockout blows, and LLM uselessness, but I find them really useful. Is there some emperor's new clothes type thing going on here - am I just a dumbass who can't see he's excited at a random noise generator?
It's like if I saw a headline about a knockout blow for cars because SomeBigBame discovered it's possible to crash them.
It wouldn't change my normal behaviour, it would just make me think "huh, I should avoid anything SomeBigName is doing with cars then if they only just realised that."
The paper shows reasoning is better than no reasoning, reasoning needs more tokens to work for simple tasks, and that models get confused when things get too complicated. Nothing interesting, on the level of what an undergrad would write for a side project. If it wasn’t “from apple” no one would be mentioning it.
Presumably his wealth would go down massively in PPP terms since if gold was so abundant in Mali back then its buying power couldn’t be that high.
In a way it’s similar to the Spanish silver mines in the New World it resulted in a significant increase in prices and a lot of wealth shifting around.
There's various possible reasons. Less pollution from tire wear and the possibility to lay tracks across grass to create relatively ecological dedicated lines so the carriages don't get stuck in traffic are two options that spring to mind immediately.
Ever is a long time. I expect first products built this exact way working reliably and having happy customers in the next five years, pessimistically. Optimistically this is probably happening somewhere as we speak.
It would be awesome to develop some theory around what kind of problems LLMs can and cannot solve. That should deter some leads pushing for solving the unsolvable with the technology.
That being said, this isn’t a knockout blow by any chance. The strength of LLMs lies in the people who are excited about them. And in this case there’s a perfect reinforcing mechanism for the excitement - the chatbots that use the models.
Admit for a second that you’re a human with biases. If you see something more frequently, you’ll think it’s more important. If you feel good when doing something, you’ll feel good about that thing. If all your friends say something, you’re likely to adopt it as your own belief.
If you have a chatbot that can talk to you more coherently than anyone you’ve ever met, and implement these two nested loops that you’ve always struggled with, you’re poised to become a fan, an enthusiast. You start to believe.
And belief is power. As in the case of neuroscience development not being able to retire the concept of the dualism of body and soul, so will the testing of LLMs not be able to retire the concept of AI poised to dominate everything soon.
DOGE staffers are widely reported to be financially independent, genius technologists from Elon's inner circle with profound lack of ethics and very narrow world perspectives that come from their extreme youth. Those guys aren't going to hang around for long and take pot shots; if they do, they'll eventually hear Trump's two favorite words anyway. The name DOGE is also politically toxic and associated with Elon Ketamine. Per your post, NYT and the rest of the liberal media will try to keep the DOGE name in the headlines as long as they can, since it is a tremendous embarrassment for Trump.
I do hope that after DOGE goes away, Trump goes back to supporting his base with respect to maintaining VA staffing, severely restricting H1B visas, etc. especially after the R&D tax credit debacle cost so many jobs and the true potential for LLMs being to increase productivity at all levels of every organization (as opposed to the false promise of halving the workforce being pitched by snake oil salesmen).
And: they have a crash cart (keyboard, mouse and display) and battery backup built-in. An old laptop is perfect for starting a homelab. The only major downside I can think of, and as another commenter already mentioned, is the limited storage (RAID) options.
Interesting piece, thanks. I also enjoyed his piece "Streetcars: An Inconvenient Truth." His argument is based on length, speed and cost; the main point is that a technically equivalent bus would often be cheaper and thus could be run on a longer, more useful route.
I think there is another aspect that usually goes unstated, which is the vibes. If you're a mayor you want to build a tram. If you're a tourist you want to ride a tram. If you're a prospective resident you want to live near a tram. Yes, it's smoother and yadda yadda, but really it's because it has more sex appeal. A technically equivalent bus may well be _technically_ equivalent but could never be sexually equivalent. Nobody would write a play entitled A Technically Equivalent Bus Named Desire. In a way, spending money on a tram is similar to spending money on parks or flowers or public art.
I wonder what the world would be like if we were honest with ourselves.
While I completely agree with you -- that's what I would want too, I want Win98 without the pain, click or type and it's there, it's a very 2025 expectation -- I laughed because of the huge disparity between this and what installing DOS and Windows were really like. Part of the experience is selecting drivers and configuration :D (Even better in DOS with the IRQs and config.sys and whatnot.)
I only experienced traditional rough sleeping homelessness once when my "house" (my van) was towed and I had to sleep in an hostile architecture bus stop bench that had ridges between each "seat" area. Otherwise, I was technically "homeless"/vanliving in SV from about 2010-2019.
I created this a while ago, and whenever I show someone they are shocked to see there is absolutely no JavaScript; all of the animations are done via marquee tags: https://udel.edu/~ianozi/
May not be the sure thing you think. A lot of (professional, pressed, retail)P Blu Rays are failing quite young. Less than a decade in some cases. Density is a bitch.
I don't think anybody who uses LLMs professionally day-to-day thinks that it can reason like human beings... If some people thought this, they fundamentally do not understand how LLMs work under the hood.
> With docker, someone has to rebuild every container that contains a copy of the library.
I think you're grossly overblowing how much work it takes to refresh your containers.
In my case, I have personal projects which have nightly builds that pull the latest version of the base image, and services are just redeployed right under your nose. All it take to do this was to add a cron trigger to the same CICD pipeline.
