LLMs don't need to be the dangerous ASI thing themselves for them to still be dangerous. It suffices if they
- destabilize the world just by what they do: making previously very hard things (generating fake everything, manipulating people, hacking, bio-weapons research) easy enough fast enough that our civilization doesn't cope with it and/or
- accelerate AI research to the point where they make the dangerous ASI thing appear years earlier than would otherwise have happened
100% this. I agree, it’s very destabilizing for humanity for the reasons you mentioned. I don’t think we know what the fallout will be yet.
Based on our past inability to handle things well, it could be extremely devastating to a large population.
I would like to think we will navigate into a new way of life to make up for the job losses and other things, but the impacts are just too devastating and we’re just getting started.
They are not going to "slow down" anyway. When has tech ever "slowed down"? The only way to "slow down" progress is by implementing anticompetitive measures, which would benefit those in lead.
A model is just another thing to plugin to a harness. I don't give it much more thought than that. If developers are still caught up on Claude Code, or Codex that's just not a long term thing. It's best to develop workflows locally and in the cloud with open harnesses. I know this will be the future because that's how it worked on every other system that developers use.
Sure there are Microslop and Oracle db users but most of the world we live in is Postgres and Linux. That's why I think most companies will run llm's like that.
That’s not true. Before AI, I have been using Jetbrains IDEs as far as I can remember. Also have been using MacBooks for work since my first job. You don’t have to generalise everything. If a particular specialised tool is good at its job just use it instead of re-inventing the wheel
> good at its job just use it instead of re-inventing the wheel
Exactly why is everyone reinventing a harness every month. There will be Microslop / Oracle harnesses and there will be 1 or 2 open source ones that win.
Even if I accepted that given the new repos: all the new code = "It's all garbage." - I would certainly love to be in a position to observe the new code + metadata = software trends, as software eats the remaining world.
My biggest problem with Zed right now is the font rendering doesn't look that good on MacOS. To me Sublime Text always had the best font rendering on Mac. Electron slowly over the years has gotten better and decent now. Zed just doesn't do the font justice. It's too thin or something.
Here we go again. Autopilot != FSD. Autopilot is not "autonomous" driving. It's lane keep with adaptive cruise control. The same system that Honda, Toyota, etc have. Yes the naming is wrong, the marketing is bad, but I don't see it as much worse as Toyota safety sense. If you use it to be "safe" you're going swerve off the highway into a ditch. I used super cruise from GM in my friends suv. As soon as lane markers go away on a bridge, I almost hit the railing.
I'll get downvoted but just giving you the facts. I'm glad the Autopilot name has been retired. Such a bad name, but maybe a good name because autopilot in planes can't see and avoid obstacles either.
How about the fact that Tesla is killing people and covering it up?
Would you go to a driver's funeral and tell their family that um, ackshully it's sparkling autopilot?
What do you think you're adding to the conversation? You're trying to distract from the fact that real, actual people have been actually killed by this.
It's not a semantic issue, FSD is a completely different system, but many people mix up the terms when discussing these systems due to poor naming. Autopilot is just cruise control and lane keep. FSD handles navigation and full vehicle control. Articles discussing the dangers of Autopilot are making perfectly reasonable claims about a system which was poorly named/marketed, but they are not meaningfully relevant to conversations about FSD.
The news isn't necessarily of the effectiveness of the particular tech stack, but the integrity, or lack thereof of the manufacturer in reporting incidents. If that is in question, assessing the effectiveness of any of Tesla's tech stacks fsd or autonomy, or taxis for driving is in doubt.
Autopilot is completely different software from FSD. If you think FSD is stupid then Autopilot is worse because it won't do anything other than stay in the same lane and adjust speed to the car in front of you.
For some reason you could turn this on when you're not driving on the highway. It doesn't do anything for traffic lights, stop signs, obstacles, etc. because it's just cruise control. It's also included with every vehicle (unlike FSD).
The difference is FSD is properly annotated as (Supervised) and does exactly that. Autopilot does not 'autopilot' the vehicle by any reasonable measure.
FSD controls every aspect of the vehicle's operation and explicitly demands human supervision. There's nothing misleading or incorrect in the name FSD (Supervised).
IMHO you're shifting goal posts (and I am not downvoting).
Tesla (or probably mostly Elon) was not selling "adaptive cruise control". It's selling "Autopilot" for $8k (now with a subscription AFAIK), with a pinky promise that "soon" or "next year" or "after two weeks" (jk) you essentially will set a destination, go to sleep and wake up at destination[1].
It's same as saying that "LLM != AI" and arguing that "ChatGPT is not AI - it's a glorified statistics model that is good at creating human sounding texts". Yeah - you and I understand this - but the average guy most likely does not and will get burned by this, because dozen tech-bros are burning billions of dollars and try to convince everyone that it's a panacea to every problem you can think of.
[1] It's a slight exageration, though I won't spend time digging for quotes but my main point is that's what Tesla are selling to an average guy and not nerds who can distinguish on what's possible, what's working and what level of driving assist there are.
"Autopilot" is not $8K, that's FSD. Autopilot was the default cruise control/lane keep software and was renamed "Traffic Aware Cruise Control" a few months ago. The original name was ridiculously misleading.
They need to lower the screening to 40. I just had mine at 40, turned out fine luckily. Did it without sedation which my doctor said was rare in US, but common outside of US. I found surprising, wasn't that big of a deal. Pain was probably at a 7/10 during the turns (like 3 times) but ok the rest of the time. A little uncomfortable. Some new sensations, some familiar (feeling like you are crapping your pants).
I walked in and walked out no issue and went on about my day. Prep was fine but would be hard if I didn't work at home.
With old school sedation I think it might be worth avoiding it. But with propofol you are out like a light, and then wake back up just as fast when they turn it off -- and it feels like you just had a nice nap. Aside from feeling a bit groggy for a few minutes, you just get up and walk out the door and go about your day. Personally, I do not think I'd volunteer for 7/10 pain just to avoid that.
My first colonoscopy was without anesthesia, and it was as described above. A little uncomfortable on the one bend, but I don't even think I would put it at 7/10 (perhaps a 4 or 5—but not a sharp pain, mind you).
Definitely enjoyed the following times with anesthesia because, of course 0/10 as far as I know. Also, anesthesia just trips my mind—how seemingly time travel (going forward in time) seems to be involved.
Sounds a bit like the transnasal esophagoscopy I opted for recently in lieu of an upper endoscopy. Thin tube through nose, then down into the esophagus. Uncomfortable while threading it through (chest pain, like swallowing a sharp tortilla chip), but ultimately not a big deal, and done in-office in about 15 minutes.
I may be getting one of those in the next few months. Just had a regular upper endoscopy for dysphagia, and they were unable to solve it or even conclusively identify a cause. They warned me that this transnasal procedure would be next. Sounds like fun...
Which state did you have it done in if I may ask? I'm in Oregon and haven't been able to find a doc that does it without sedation. I can't be put under sedation for medical reasons, but I definitely need this procedure done sooner than later due to new GI problems.
I had a colonoscopy and esophagogastroduodenoscopy (thankfully not in that order) without sedation. Had it done in D.C.
My doc looked at me like I was crazy when I asked if it could be done without sedation, and reminded me that it would be uncomfortable, but otherwise didn't have any problem with it. I've endured 50k runs, brutal workouts, and traumatizing childhood neglect - I really can't see what the fuss is with mild discomfort that, by comparison, barely registers, and for such a short amount of time at that.
I too had esophagogastroduodenoscopy and the "sedation" I received as a barely noticeable dose of fentanyl. It was unpleasant to feel like I was drowning in saliva but it was quite bearable.
If I ever receive that procedure again, I will ask to skip the fentanyl microdose. The anesthesia and the buzz were not only underwhelming but for some reasons I started to feel the typical opioid warmth when the procedure was almost completed. Had they waited a few minutes after the IM injection I might have had another opinion on the usefulness on fentanyl during endoscopie because the last 30s were almost pleasant!
I don't really mind the pain itself, but I could see myself thinking the worst in that situation and imagining the strong pain meant the probe had punctured my colon.
I assume that's not actually a realistic risk, right?
North Carolina. And I wasn't actually aware that some doctors wouldn't do until after I had it done (reading about it online). I just called before my appointment to say I didn't need sedation. They said ok and wrote it down. They weren't really pushy during the appointment other than asking me why I didn't want sedation.
I thought it was going to be awkward but wasn't at all. We just chatted. It was him and an assistant. I was able to watch the TV of my colon while he was doing it.
For people who might be interested in following your advice, the conventional wisdom is that you should definitely look around when choosing the doctor. I.e. do not use a regular gastroenterologist who primarily does sedated colonoscopies, you want one who has experience with non-sedated ones. They have a better idea of what hurts and how to mitigate that.
Do you have specific risk factors that caused your doctor to recommend getting it at 40, or did you have to convince them? My understanding is that if the doctor doesn't order it, many insurance companies won't cover it.
I had bleeding, light red, most likely from hemorrhoids (it was), but had a few other factors that the doctor took it seriously. The insurance didn't cover much as far as I know (or any?). Owed like $1800.
There's an option which doesn't involve drinking any yucky fluids, just water. SuTab. You have two rounds of twelve pills that you drink with three cups of water at various intervals.
I've done it enough times that I'm totally fine with it: the electrolyte drink tastes like slightly bad-tasting Gatorade, which is hardly worth caring about.
And you get diarrhea-like bathroom runs half a dozen times maybe.
Yes it was annoying to get the runs and gross to drink the stuff the first few times, but people eat things like cow tongue or live octopus or whatever... I can handle some bad-tasting Gatorade and some diarrhea just fine, especially given the 5 years of peace-of-mind it buys me afterward.
People react differently. I took the same prep but the waves of nausea and cramping were so intensely painful I sweat through all my clothes and passed out on the toilet. It was some of the worst hours of my life. I’m just over 40 and have some symptoms to check which turned out to be benign, but it was such a harrowing experience I will be doing this as infrequently as possible.
Same thing happened to a family member. Dunno why the effect is so variable- I guessed it was more of a mental / expectations thing initially than a direct physical response. I see now I was probably wrong.
I had a colonoscopy + upper endocoscopy a few months ago, age 29.
The prep was horrible, particularly the electrolyte drink they make you take the night before. I almost puked several times trying to get that stuff down.
Actual procedure was a breeze. I was sedated, and then I woke up and it was over.
"I can state flatly that heavier than air flying machines are impossible.
— Lord Kelvin, 1895"
I'm sure this doesn't apply to you since you're not Lord Kelvin. On the other hand, people like Peter Norvig state in a popular AI textbook that, for example, they don't know why similar concepts appear close by in the vector space, so maybe you just know something other people don't.
I'm not taking a position here but the person you're replying to stated that Anthropic are working on AGI, not that their current LLM offering will evolve into AGI.
False, and you've given no argument to the contrary. There's certainly no definition that precludes it. It isn't, currently; there's no reason it can't be, any more than there's reason that Conway's Game of Life can't be, given sufficiently interesting data to process. Any Turing-complete system could simulate AGI. It might not be the most efficient mechanism for doing so, but that's not the question at hand.
At worst, this would only be temporary. Yes, it will cause hardship, but once it trickles down to regular companies, that an LLM is a word predictor, workers will be back. Seems Andrew Yang has never used an LLM to the lengths we do as software developers.
The thing we have to worry about is what's after the LLM.
I don't understand why "it's just predicting words, bro" is still seen as a valuable argument. A LOT has to happen to accurately predict the next word(s) for any given topic.
If that supposed to be a dismissal, it's not a good one.
reply