8,956 karma · joined March 13, 2013
If this has been done appropriately and safely, it should be similar.
One thing I will mention, having watched the video, is that I think this goes about the same speed as what the pros do? It takes around 20 seconds for removal of the whole hair (about 10 seconds of current plus 10 seconds tweezing, inserting needle, etc.) in the video. My tech can usually remove around 200 hairs in an hour, which comes out to about the same (200 / 60 = 3.33 hairs per minute = around 18 seconds a hair).
It's crazy how even something which feels mediocre so much of the time - fast-food coffee, a budget airline - requires an enormous amount of human effort to pull off reliably.
(And yes, you can dislike Southwest as a corporation and still think things like flight attendant training and plane simulators are cool. Come on folks.)
But it's no way to live and we're seeing the consequences of it now.
To be clear, I'm not asking if you're against DEI and support a review on those grounds (I'd also disagree, but that's whatever). I'm asking, do you think that's really why they're doing this, and if it isn't, do you support the actual reasons?
I see that folks could see this as evidence that I live in a bubble, but I genuinely cannot think of a way this is not an authoritarian move.
My wife is a physician, and when permitted by patients, uses one of these tools. It's been an enormous time-saver for her. She works a 32-hour week, meaning 32 hours of seeing patients. Before these tools, she was regularly spending an extra 8 to 16 hours, e. g. two full work days, writing notes and sending messages. That time has been more than cut in half. She would never give up the tool if she didn't have a choice.
According to her, it is reasonably accurate, but all notes must be manually reviewed (not just as in her organization requires that, but also as in if she didn't, it would be obvious due to its mistakes). The biggest issue is with things like names and medications, stuff that isn't present in ordinary English, as well as mishearing the results of diagnostic tests, numbers, etc.
It's rare for patients to refuse it.
If you are capable of being hired at Palantir, there are thousands of non-evil companies who would be happy to have you.
Most of my machines come with at least some defects. Some are ready to go in a few minutes, some take several hours or days. I mainly do it because I enjoy it. I try to not lose money on parts or other inputs, besides my time.
As of today, I've sold more than 800 machines at an average of $80 per machine and an average profit of $30 (approximately). That's around $24,000 of profit over the last three years or so. It covers all the costs of its own inventory, parts, losses (e. g. some machines just never make it to sale), and it's built a lot of fun community relationships. Plus, I've helped a lot of people get access to a working computer at a low cost!
This would never, ever scale beyond me doing it. The moment I had to employ a person, pay rent on a space, or start offering warranties and free returns and so on, that profit margin would vanish. That's why it's a hobby, not a full-time job. I do it on nights, weekends, and in between working my day job (e. g. I'll have a Windows install going in the background while I code).
But it's fun, it's valuable, I've learned a lot about running a business, and it's paid my car payment a few times. It's also nice to have a 'job' that is very different from my day job: much more hands on, not as much complex thinking required, and more immediately rewarding. (At least for me. I just love when a broken thing starts to work again.) The hardest part is the customers, especially when things don't go well (e. g. are my fault) or they are in a bad mood.
I think more people should do things like this. It doesn't have to be the thing that gives you the money you live on to be valuable.
It's not enough to offer incentives. You have to explain why you even need to do it in the first place, and the answer better not be "we want money."
What, ah, do they have to do with AI?
First, something lying sometimes doesn't mean it lies all the time.
Second, the whole point of LLMs is inference - they use massive amounts of amalgamated information to produce answers. The Olympiad math problems are not frontier mathematics requiring ideation, they are complex examples of existing problems. That means they're exactly the sort of thing an LLM with enough training data is good at.
The question of whether recombining existing knowledge is all it takes to be "creative" or produce things which are novel is an open one, but I don't think this is contradictory on its face.
It might have involved OP getting a taxi to the airport and being chased down at the gate, but it would have worked.
And even if it didn't, they would have saved themselves enormous time, money, and stress by following through on their threats if it didn't work out.
But that said, it took more than forty years to electrify the entire United States[0]. "The internet", as we think of it, hasn't even existed for 35 years yet. (Yes, I know there were networks before that that the current system arose from, but that's hair-splitting. I don't think the kind of Internet the average person might even consider using existed before, generously, ~1995-1996.) Yet, 95% of US adults use the Internet, implying a penetration at least as high[1].
The median Internet speed in the US is around 250mbps down and is in the top 10 in the world[3].
The problem is that access and speeds are not evenly distributed, not that we can't get 22gbps symmetric down/up. We don't need to give people in cities faster Internet; truly, you do not need that speed to do day-to-day tasks. You don't even need the 1gbps down/150mbps up that I have. What we need to do is make sure people in rural areas can access at least the median speed.
That said, I think we could give it another 15-20 years and see where our country with around 36 times the population and 238 times the landmass is at in terms of speeds.
[0] https://www.christenseninstitute.org/blog/how-the-history-of... [1] https://www.pewresearch.org/internet/fact-sheet/internet-bro... [2] https://tachus.com/internet-speeds-usa-vs-the-rest-of-the-wo...
Edit: For those who are curious, I took this test: https://www.typing.com/student/typing-test/1-minute and got 116 wpm at 99% accuracy.
At https://www.typingtest.com/benchmark.php over two minutes, 117 wpm at 97% accuracy.
It does mean that I can't use "ergonomic" or split hand keyboards easily, though.
I have a couple different inexpensive Aula and similar mechanical keyboards I prefer to use.
Meanwhile, the cardboard box you have forgotten to take to the recycling for three weeks will become their palace.
That aside, I simultaneously hold several thoughts about this.
The first is: The TSA probably doesn't need to exist at all in its current form; I'm not aware of any evidence they actually do anything to stop actual threats.
The second is: I have sympathy for the folks who work there, who probably are in it mostly to have a stable and non-controversial government job. I'm sure they're going through it, for political reasons that are not their fault.
The third is: There is zero political incentive for either side to cave on this, that I can see, and as somebody who is generally opposed to the behavior of ICE, I think we as the American public will just have to eat it until they are reined in. I'm including myself in that; this has messed with my travel plans and will continue to do so.
But if I were on the other side of the issue, I think I'd hold the exact same opinion: There is no incentive not to continue the shutdown.
I think the reality is that Epic got big because of Fortnite but nothing lasts forever. They would have been better off building a war chest and pulling a Valve (though I'm sure they'd hate hearing it that way): going silent and making whatever they wanted for a while, and then trying to repeat the cycle, rebuilding the chest, and then going on. Video games are the exact opposite of Infinite Growth Forever. People get bored and move on.
Meanwhile, Epic has many stable and valuable businesses - Unreal, the game store, etc. - which are perfectly capable of sustaining a sizable company. Just not one as sizable as Epic is. The best case for them is they figure that out, and manage to make a sustainable go of it doing that.
Still, this is a great idea, and one I hope takes off. I think there's a good argument that the future of AI is in locally-trained models for everyone, rather than relying on a big company's own model.
One thought: The ability to conveniently get this onto a 240v circuit would be nice. Having to find two different 120v circuits to plug this into will be a pain for many folks.
What I find most incredible about it is the number of women who experienced immensely painful procedures while conscious multiple times, and went back again and again, in order to have a child. And few of them, if any, regret it.
It is simultaneously one of the most impressive feats of modern science, and one of the most unfair burdens put on any section of the populace, that they were able to, and had to.
On a pure neurological basis, this is just untrue. It's well-established in research that the brain is massively parallel and one of the main differences between it and a digital computer is that it is doing a lot of different things all at once, not switching between different things quickly.
Here are some references:
[0] https://psychology.stackexchange.com/questions/1946/differen... (good lay explanation) [1] https://pmc.ncbi.nlm.nih.gov/articles/PMC4387515/ [2] https://www.sciencedirect.com/topics/neuroscience/parallel-p... [3] https://esmed.org/parallel-processing-in-problem-solving-a-n...
(For the papers, I'm not necessarily advocating for their conclusions, but they're a good jumping-off point to see all the things they cite, which shows a consensus about how the brain works neurologically.)
I suppose that doesn't mean that our conscious cognition must also be parallel, but there's not really a lot of good reason to assume it _wouldn't_ be, other than that it fits Chater's position.