Also players don’t know what they want. Good games aren’t just a result of a good idea that’s then implemented - they come from untold hours of iteration, tinkering, figuring out what’s fun, what isn’t, and why. That’s the hard part, and I have a hard time believing that hypothetical holodeck could ever do it.
There are large datasets of bird sounds (eBird, Cornell Labs, etc), but the descriptions are usually limited to the species, location, and something like “mating call” or “contact call”. Hugely useful for building models that can recognise birds by call (apps like that already exist, I recommend Merlin Bird ID), but definitely not enough for something approaching actual translation.
FWIW there’s no research to indicate that the sounds birds make are what we’d call language. They’re avid communicators, and some species are known to be highly intelligent, but of course it’s not like “caw” means “to fly” and “craah” is “forward”.
I’ve been winter feeding the local birds for a couple years now, and especially the great tits acknowledge my presence via a special call I haven’t heard anywhere else. It’s distinct from the calls they use when they find food otherwise. Bird ID applications consistently fail to recognize the species based on that call.
There are stationary feeders in the neighborhood, but the birds don’t seem to associate the humans filling them with their food, and subsequently just use warning calls when they see humans approaching the feeders.
But whenever they see me, even in the summer, they use that call. Blue tits have their own, shorter variation of it.
I like to think they’ve given me a name in their language :-)
In my experience LLMs are surprisingly bad at CSS beyond a very basic level. They work fine if you need to change the color of a button, but when it comes to actual styling work, even intermediate stuff like position:absolute or CSS grid, Copilot or even CC default to outputting correct-looking gibberish really quickly.
That's telling about CSS design. Folks here on HN are talking about how they purposely ask LLMs about APIs that don't exist, and they hallucinate with a better and more intuitive design that they would come up with on their own.
I don't know the best solution for the problem, but CSS is a very convoluted one.
My guess is it’s because CSS is so dependent on context. Especially layout styles only make sense for a specific structure of HTML elements, which might be stored in an entirely different file and directory.
“Somewhere in the mix”? Sure. I’ve seen TTRPG folks experimenting with LLMs for about as long as LLMs have been around. A couple years later it still can’t really run a session, but you might find it useful for generating some details of an adventure.
The big issue really is making its output interesting. As usual, an LLM will default to the most generic outcomes and settings. For a TTRPG to be fun you usually want surprise, drama, creativity - it can’t do any of those.
> We made approximately the same progress in the past 100 years as the prior 1000 as the prior 30,000
I hear this sort of argument all the time, but what is it even based on? There’s no clear definition of scientific and technological progress, much less something that’s measurable clearly enough to make claims like this.
As I understand it, the idea is simply “Ooo, look, it took ten thousand years to go from fire to wheel, but only a couple hundred to go from printing press to airplane!!!”, and I guess that’s true (at least if you have a very juvenile, Sid Meier’s Civilization-like understanding of what history even is) but it’s also nonsense to try and extrapolate actual numbers from it.
Plotting the highest observable assembly index over time will yield an exponential curve starting from the beginning of the universe. This is the closest I’m aware of to a mathematical model quantifying the distinct impression that local complexity has been increasing exponentially.
It’s plausible in a sci-fi sort of way, but where does the model come from? After a hundred years of focused study we’re kinda beginning to understand what’s going on inside a fruit fly, how are we going to provide the machine with “a perfect model of all interactions between biological/chemical processes”?
If you had that perfect model, you’ve basically solved an entire field of science. There wouldn’t be a lot more to learn by plugging it into a computer afterwards.
Sure, let’s send private medical data to a cloud server somewhere for processing, because a medical professional in 2025 can’t be expected to know how to use a keyboard. That’s absurd.
I can type quite well. I can also troubleshoot minor IT issues. Neither is a better use of my time than seeing patients.
I’m in an unusual situation as an anesthesiologist; I don’t have a clinic to worry about, so my rate-limiting factor isn’t me, it’s the surgeon. EMR is extremely helpful for me because 90% of my preop workup is based on documentation, and EMR not only makes that easy but lets me do it while I still have the previous patient under anesthesia. I actually need to talk to 95% of patients for about 30 seconds, no more.
But my wife is primarily a thinking rather than doing doctor, and while she can type well, why in the hell do we want doctors being typists for dictation of their exams? Yes, back in the old days, doctors did it by hand, but they also wrote things like “exam normal” for a well-baby visit. You can’t get paid for that today; you have to generate a couple of paragraphs that say “exam normal”.
Incidentally, as for cloud service, if your hospital uses Epic, your patients’ info is already shared, so security is already out of your hands.
Macs have pretty decent on-device transcription these days. That’s what I set up for my wife and her practice’s owner for dictation because a whole lot of privacy issues disappear with that setup.
The absurdity is that doctors have to enter a metric shit ton of information after every single visit even when there’s no clearly compelling need for it for simple office visits beyond “insurance and/or Medicare” requires it. If you’re being seen for the first time because of chest pain, sure. If you’re returning for a follow up for a laceration you had sewn closed, “patient is in similar condition as last time, but the wound has healed and left a small scar” would be medically sufficient. Alas, no, the doctor still has to dictate “Crime and Punishment” to get paid.
This has been happening for years, long pre-dating LLMs or the current AI hype. There are a huge number of companies in the medical transcription space.
Some are software companies that ingest data to the cloud as you say. Some are remote/phone transcription services, which pass voice data to humans to transcribe it. Those humans then store it in the cloud when it is returned to a doctor's office. Some are EMR-integrated transcription services which are either cloud-based with the rest of the EMR or, for on-premise EMRs, ship data to/from the cloud for transcription.
Medical companies could self host their speech to text translation. At the end the medical data is also on some servers stored. So doing speech -> text translation seems just efficient and not too much worrying if done properly.
So you think the better solution to doctors not being able to try is for them to self-host a speech to text translation systems, rather than teaching doctors to type faster?
Their healthcare/IT provider like Epic would do it. And in fact some have already done it, from what I can see.
Furthermore, preparing/capturing docs is just one type of task specialization and isn’t that crazy: stenographers in courtrooms or historically secretaries taking dictation come to mind. Should we throw away an otherwise perfectly good doctor just for typing skills?
I imagine where the speech to text listens to the final diagnosis (or even the consultation) and summarizes everything in a PDF. Of course privacy aware (maybe some local hosted form).
And then the doctors double checks and signs everything.
I feel like, often you go to the doctor an 80% of the time they stare at the screen and type something. If this could get automated and more time is spent on the patient, great!
Who is responsible when the speech-to-text model (which often works well, but isn’t trained on the thousands of similar-sounding drug names) prescribes Klonopin instead of Clonidine and the patient ends up in a coma?
This isn't a speech recognition problem per se. The attending physician is legally accountable regardless of who does the transcription. Human transcriptionists also make mistakes. That's why physicians are required to sign the report before it becomes a final part of the patient chart.
In a lot of provider organizations, certain doctors are chronically late about reviewing and signing their reports. This slows down the revenue cycle because bills can't be sent out without final documentation so the administrative staff have to nag the doctors to clear their backlogs.
You’d be surprised. Point-and-shoot cameras have become extremely popular with young people in the past ~2 years or so because of the nostalgia factor.
I see people using weird things like 2000s digital cameras and Nintendo DS cameras for that old look, but I've never seen someone with one of those entry to mid level point and shoot cameras you used to see before smartphones. I only see phones, ancient retro cameras, and hobbyists with high end gear.
