I wonder if AI also reveals unnecessary parts of the workforce by demonstrating that what they do is actually pretty trivial.
There are a ton of basically BS office jobs that could probably be replaced by AI, or in some cases just revealed as superfluous.
We need to just stop pretending we still need a 1:1 connection between employment and income and do UBI. Useless jobs help us preserve the illusions of a pre-post-industrial civilization. Instead of just paying people, we pay people to do work we don't need.
The joke about someone using chatGPT to write a lengthy email that the recipient will summarize with ChatGPT is the perfect example of how pretend much work is.
There is this joke about socialism where hundreds of workers digging with shovels and somebody asks “Why not use that excavator? One machine could do it in no time” and the other answers “And put 20 men out of work? We’re creating jobs!”.
Still far, far too complex to occur "randomly," which is fascinating. The odds of 112k bases arranging in any meaningful way by chance within a membrane are the kind of thing you wouldn't get if you ran a trillion trillion trillion universes.
There's many hypotheses, basically all different variations on "soup of organic compounds forming complex catalytic cycles that eventually result in the soup producing more similar soup, at which point it begins to be subject to differential selection." It's a reasonable idea but where did this happen, and do the conditions still exist? If we went to that place would it still be happening?
There's reason to believe the answer would be no because modern lifeforms would probably find this goo nutritious. So life may have chemically pulled up the ladder from itself once it formed.
This of course assumes no to more fanciful options: panspermia that pushes the origin back to the beginning of the cosmos and gives you more billions of years, creation by a God or some other kind of supernatural or extra-dimensional entity, etc.
1. Autocatalytic RNA reaction networks -- "soup producing more soup" -- are easily replicated in the lab, subject to Darwinean processes, and are at the center of ongoing study. "0 to Darwin" is now easy, "Darwin to Life" is the new focus, and God of the Gaps must retreat once again.
2. Spores hitchhiking on impact ejecta sounds exotic until you realize that anywhere life is present at all spores will be everywhere and extremely sturdy. That desktop wallpaper you have of planets crashing together and kicking off an epic debris cloud? Everything not molten is full of spores.
3. Religious explanations are not in the same universe of seriousness as 1 and 2. Opening with a religious talking point and closing with a false equivalence is mega sus.
Not to mention the constant trickle of "X survived in space" stories that we get every time someone bothers to collect and culture a sample. The amount of success at every stage with, frankly, very little effort spent tuning the conditions, multiplied by "bacteria are everywhere" makes hitchhiking less crazy than it sounds. Our intuition misleads us because bacteria are so much better at handling acceleration (easy if you're small) and dessiccation (everywhere is a desert if you're small) than anything we are used to thinking about.
Good news: the primordial oceans were so vast (literally planet-scale) and persisted for so long (millions to billions of years) that you can run a trillion trillion individual random reactions.
You are being severely restricted by your imagination. You seem to have presupposed that random abiogenesis is impossible and reconstructed the facts to support that claim because you can't conceive of the alternative.
Planets are really, really big. Any one chemical reaction is on the scale of molecules. If you let those figures compound for a long time, the number of total reactions gets very, very large. Far larger than you imagine. Many times more.
>Still far, far too complex to occur "randomly," which is fascinating
I don't see the word "random" anywhere in the article. By random maybe you mean it's seemingly indeterministic? Regardless of the nature of the underlying process, at the classical level, the environment acts as a deterministic filter, ie, other chemical processes.
> Still far, far too complex to occur "randomly," which is fascinating
Why spend time making this point? Nobody believes that this occurred randomly: it occurred via evolution.
The mutations are a random part of evolution, but the process overall is not random at all - no more so than your immune system (which randomly generates antibodies, then selects against those that target innate epitopes), or stable diffusion (which starts with random noise, then marches up a gradient toward a known target).
It is the selection step that makes similar processes non-random, because a random selection step would just be noise.
This is technically random. The entire creationist argument is that complexity cannot come from randomness but evolution is the method in which it does.
Evolution is just a sort of way for low entropy structures to form from randomness. It’s still random all the way down.
The man is just trying to reconcile a belief in god with the scientific reality. He needs to bend the evidence to fit his identity he cannot bend his identity to fit the evidence because that could break his identity. The fact he commented here on this topic is sort of unhinged. It seems like the article presented evidence that is strikingly against his world view and he needed to justify something in order to prevent his identity from rearranging itself according to external reality.
If you’re interested in this area I highly recommend “The Vital Question” by Nick Lane if you haven’t read it.
The TLDR of his theory is that life originated in alkaline hydrothermal vents on the ocean floor, where natural energy gradients could have driven primitive metabolic reactions before the development of DNA.
Book goes into a lot of layperson-accessible detail.
I’d recommend talking to people in the trades first. Not saying it can’t be a good move, but it is definitely hard and has its own huge downsides like poor working environments, long hours, and years to actually get into decent paying roles.
Faddish career advice is usually bullshit, or it’s too late and the bus it’s telling you to jump on left the stop years ago.
It’s mostly ZIRP hangover. The over hiring in those years was ludicrous.
AI is having some effect at the margins but it’s mostly an excuse.
Companies always prefer to avoid saying they are just laying people off. It can be a negative market signal to investors, which is paradoxical, since it might indicate lower growth expectations. It also creates possible exposure to lawsuits depending on the circumstances and state.
The nice thing about AI as an excuse is you can say to your investors and board “we are shedding cost but still growing and in fact our productivity is up because we are leveraging AI!”
Can't find 4.5, but 3.5 Sonnet is apparently about 175 billion parameters. At 8-bit quantization that would fit on a box with 192 gigs of unified RAM.
The most RAM you can currently get in a MacBook is 128 gigs, I think, and that's a pricey machine, but it could run such a model at 4-bit or 5-bit quantization.
As time goes on it only gets cheaper, so yes this is possible.
The question is whether bigger and bigger models will keep getting better. What I'm seeing suggests we will see a plateau, so probably not forever. Eventually affordable endpoint hardware will catch up.
If humans were much more rational this would work better.
The human brain is loaded with exploits, and capitalism being an excellent optimizer quickly finds and uses these exploits. Because they work, and more importantly they are way way easier than creating actual value.
A casino is more profitable than a hospital. Quack medicine sold with sensationalism is more profitable than real medicine. Porn is more profitable than good film or literature. Rage inducing click bait is more profitable than actual news or thoughtful editorial. It’s kind of just thermodynamics. These things require less energy input, and they don’t have to “work” because they exploit security vulnerabilities in the dopamine system instead.
I do legit wonder about K abuse in this whole crowd.
Anyone hear any excerpts from Thiel’s Antichrist lectures? I’ve never been on the same page as him politically but this wasn’t “wow I really disagree” material. This was “are you… okay, man?” material. It was just askew and bizarre. I’m not a big Thunberg fan either but I cannot replicate the thought process that would lead to mentioning her as a potential Antichrist prototype. And that wasn’t even the weirdest thing, just the easiest to explain.
One thing I learned way back in college: if someone seems like they are on drugs, they may be on drugs.
Either that or these guys are in some weird echo chambers.
> I’m not a big Thunberg fan either but I cannot replicate the thought process that would lead to mentioning her as a potential Antichrist prototype. And that wasn’t even the weirdest thing, just the easiest to explain.
I know he doesn't like Greta, I don't either.
But I didn't see his lecture, he theorizes that she may be the antichrist? lol
I’ve come to see this as a general rule: trash maximizes engagement.
By trash I basically mean either porn or gambling. By porn I don’t just mean the sexual kind but also political rage porn, etc. By gambling I mean anything that exploits the kinds of dopamine hooks that a slot machine exploits. There are many variations of these things but those are the basic forms.
Those are the kinds of things you get if you optimize for engagement.
You also get more predators and trolls because those are the kinds of people who create the most engaging content.
This isn’t new. It’s been known since mass media was invented. “If it bleeds it leads,” the P.T. Barnum principle of “any publicity is good publicity,” and so on.
What I think is new is the degree of individualized hyper optimization two way digital platforms allow. They let us turn this so far up that apps on a little pocket computer can start rivaling cigarettes for addictive qualities and psychological harm.
The EV1 was way, way ahead of its time, and was more or less outright killed for various reasons including car makers having deep sunk cost in ICE engine tech. The battery tech back then was vastly inferior to today but it was still good enough for a shorter-range economy EV that could have replaced a gas for a lot of daily commuter drives, especially for two-car families.
For reference: the first generation Nissan Leaf had similar range to the EV1. I still have one of these. It's our family's second car, and has run flawlessly for over 10 years with virtually zero maintenance. Range is still about 60 miles per charge.
BTW... despite the antics of Musk, I think he was absolutely instrumental in advancing car electrification. Yes there were others making EVs, but Tesla was the first to make them cool and in so doing force the rest of the industry to move. Without Tesla dragging the industry kicking and screaming into EVs I think we'd still be stuck with almost 100% ICE cars. China might have done it, but that's because they don't have the same sunk cost in ICE engines we have.
Pre-Tesla EV companies were kind of stuck in a catch-22, where they couldn't enter the low end of the market (because the tech was still super costly) and they couldn't enter the high end of the market (because either the performance was lacking or they didn't have the resources to scale production).
But with a combination of throwing a lot of money at the problem, being in the right place at the right time, and good execution, they managed to scale up the high end of the market enough to eventually move (somewhat) down-market.
Why did you name check Musk and not the actual founders of Tesla, Eberhard and Tarpenning, that took advantage of their experience with lithium batteries and the forward-thinking Californian regulatory regime to found their company there to build an electric sports car?
They didn't found their company to build electric sports cars, they founded their company to design and sell conversion kits, and famously got into a big argument with Musk when he wanted to sell complete cars.
Good point, and they're worth name checking too, but the company didn't go anywhere big until Musk took over.
Musk clearly has (or at least had) a great skill when it comes to scaling companies doing hard things. If he had one under his belt, like Tesla, I'd be willing to chalk it up to luck, but he has two: Tesla and SpaceX. Both have been spectacular successes doing things most people run away from screaming with their tail between their legs, namely volume production of innovative cars and aerospace.
IMHO SpaceX is a lot more impressive. There's an old joke: how do you become a millionaire? A: start as a billionaire and found an aerospace company.
I think without him Tesla would have been a boutique car company. They would have made expensive boutique cars for a subset of visionary EV early adopters, but would not have moved the industry. To move the industry you have to grow hard and fast enough to scare legacy car makers into trying to play catch-up, which is what Tesla did. The only other thing I can imagine moving big car makers like this would have been the government mandating an EV transition. Big bureaucratic things only move when kicked.
People hate acknowledging this because Musk's politics have turned so many people off, but unfortunately there appears to be no correlation between skills in one domain and being a generally well adjusted human being. The world doesn't work this way. A person can be good at something and still be a lunatic or an asshole.
I mean... if we dismissed all achievement of people who were assholes or lunatics, we'd basically have to throw out 2/3 of all music.
If anything there might be a slight negative correlation between extreme skill in some domain and being well adjusted, for a variety of reasons including the weird way people treat "savants" in any field. I also suspect a big one is that people with messed up backgrounds (bad childhood, etc.) or psychological issues sometimes "over-compensate" by achieving hyper-skill in some area.
I see Musk in the the same way that I see Steve Jobs. Steve Jobs was famous for his "reality distortion field".
I think that Musk/Tesla gave Politicians the idea that everyone can go electric and as a result it ended up being mandated into law. Then manufacturers had to try to make cars which were electric.
I'm not sure that's all it was, but him not being invited was definitely a travesty and clearly a result of in-crowd lobbying from established car makers with links to Washington.
I had a couple of coworkers who had them. I vaguely remember it being based on a ~$35k value in 1997 dollars, so definitely out of my price range as a new graduate earning only a little more than that in a year.
On top of that I'm pretty sure the unit economics were firmly in the negative, even discounting the R&D costs.
They were pretty remarkable though—I got a chance to drive a pre-production one at a ride-and-drive a year before and was super impressed.
There are a ton of basically BS office jobs that could probably be replaced by AI, or in some cases just revealed as superfluous.
We need to just stop pretending we still need a 1:1 connection between employment and income and do UBI. Useless jobs help us preserve the illusions of a pre-post-industrial civilization. Instead of just paying people, we pay people to do work we don't need.
reply