monotremata

Agreed.

That's still about 1000x less raw CPU performance than the cheapest smartphone on the market, though.

(E.g. a Blu View 5 Pro, which is the cheapest Tracfone is listing as in stock, for $30. Its Helio G36 uses 4 Cortex A53 cores at 2200MHz and 4 at 1800 MHz, and between 1-2 instructions per clock, giving between 16000-32000 million instructions per second. The IBM ran at 16.6 million instructions per second, per that Wikipedia link.)

(Looking at Walmart turned up a TCL K33 5G for only $20, but it uses a Dimensity 6300 cpu which is actually a little more powerful.)

And indeed figuring out what problem you're actually supposed to solve. People assume they can just give an LLM the same spec they'd give a programmer, and that doesn't work, because the programmer will figure out all the hidden shit they omitted from the spec but which is actually absolutely critical to its function. They'll also anticipate future refinements of specs and how to leave room to accommodate those changes. Etc.

I really feel like the author of this piece got this strawman in their head and didn't bother looking into how people saying these things are actually using them, because most of the people I hear saying them are themselves programmers, and they're not trying to say "my job is trivial." They're saying "the thing the LLM does fast is not my whole job. it leaves out all the important stuff."

Yeah! Same kind of thing. My sister's kid would do it also for fairly minor things that had just become frustrating, like if they couldn't get a container open, or the clothes wouldn't fit on the Calico Critter, etc. They'd try, but then as they got frustrated, before they reached the point of actually crying about it, they would bring it over and present it to Dada and say "oh no wha happen!" And he knew he should give them his full attention as if they'd started crying, and comfort them in addition to solving the direct problem. It wasn't grammatically sophisticated, but it was very effective communication.

I'd like to point out that children do, in fact, get confused in this way.

When my sister's child (who now uses they/them pronouns) was quite young (maybe three years old? I'm not certain), when something upset them, they would say "oh no wha happen?" because they knew that when their parents responded with concern after they scraped a knee or something, they would always respond with "Oh no! What happened?" So even though the new situation didn't rise to the level of provoking crying or whatever, because they wanted that same level of attention and concern, they just directly asked for that response. It was pretty cute.

Similarly, when they wanted something to happen, they would ask it as a question: "want to make those guys talk?" (which meant "please make believe with me and provide voices for some of these stuffed animals"). Or "want to just bring the bowl down?" (which meant "I'm not sure which candy I would like! please bring the candy dish down so I can pick them by looking at them [and maybe take more than one]"). So they were definitely imitating things their parents said before understanding the full grammar of the situations involved.

I mean, I don't have the encyclopedia anymore, so I guess I can't really prove it to you. But it was sort of my introduction to the idea that not everything in an encyclopedia is necessarily true, so it's pretty memorable for me. I recognize the irony of using another encyclopedia here, but Wikipedia mentions in the History section of its Pain in Animals article, "Researchers remained unsure into the 1980s as to whether animals experience pain." (https://en.wikipedia.org/wiki/Pain_in_animals#History ) I obviously don't know the story about how the line ended up in World Book, but I know they employ fact checkers for every line in the book, so presumably in the 50's the line reflected some kind of scientific consensus, as bizarre as that sounds now.

When I was a kid in the 80's, my sister and I found an old World Book Encyclopedia from the 50's that someone in our neighborhood was throwing out, so we brought it home. It had some interesting "facts" in it. E.g., did you know that guinea pigs are perfect as subjects for scientific research because they're unable to feel pain? 1950's worldbook did!

Earlier tonight it also wasn't showing up for me in search results on Duck Duck Go. All the reddit links were just labeled "reddit," and the text was "We would like to show you a description here but the site won't allow us." Weirdly it doesn't seem to do it if "reddit" is one is the search terms.

If this is part of their strategy going forward, though, that's gonna hurt their clickthrough too.

Probably a Funkey? There have been various others (PocketSprite?). Anything smaller than a Miyoo Mini is really too small to practically play on, though. And even that is pushing it for my aging eyes.

They ported Half-Life to PS2. And the sequel was on everything.

Yeah. It certainly pays off sometimes. Amazon did it. It just, y'know, also crashes and burns sometimes, and I'm not sanguine about the way this is shifting its investment money from venture capitalists to, y'know, passive index fund investors.

No, my understanding is that they're bringing in revenue on token generation, but it's exceeded by the costs of token generation (running data centers, so, electricity and cooling). They definitely want to make a profit on token generation, but they're afraid that raising costs that high too quickly would drive customers to switch to other providers. So they've reduced the amount they're subsidizing token costs, but not switched over to making a profit.

I can't find a good citation for this, though, so it's possible I'm mistaken. They also have huge costs associated with buying new GPUs and building new datacenters, so they're operating at a massive loss either way, and it's a little hard to find articles which tease apart the two aspects.

In any case, operating at a massive loss for the first few years is practically standard operating procedure in silicon valley at this point, and sometimes it eventually leads to a profitable, even wildly profitable, business (e.g. Amazon). But it does require a steady stream of investors and a steadily increasing market valuation. That's...we'll have to see what happens on that front.