HN user
I'm thinking that over the course of the next 10-18 months the price should continue to come down as pre-IPO shareholders become unlocked and can sell their shares.
All these bans on building data centers on Earth are starting to make this wild proposition seem a little more rational.
It's a scam only if you think the stock market in general is a scam. Remember, SpaceX is completely globally dominant in its industries (launch technology and orbital telecom)
It's right at IPO price, which (thus far) means it was more or less priced correctly.
It's exactly the opposite as you say though, no? Canada has much more market concentration than the US, probably a major cause of the higher prices: https://alphabridge.co/economy/canadas-oligopoly-one-big-cou...
Germany has had an immense amount of immigration over the past couple decades.
Not me. I have a Mistel Barocco and am still at 100WPM
Some people have a consensus-based mechanism for identifying if something is true or moral or fashionable. They basically believe whatever is the majority opinion, weighted for perceived social status or expertise or whatever else. I think this is normal for most non-technical people who don't have strong critical thinking skills themselves. So they outsource all the heavy thinking to the society at large.
In the end it's really just greed.
Yeah, well their job is to get the best price for their product. Just as it is your job to get as much money as possible for your product, i.e. your talents and labors.
In a competitive economy with informed buyers this greed is what makes things cheap and high quality. Think about two grocery store owners in a small town. They've settled into an equilibria with each other to keep the status quo and not get greedy for a bigger share of the market by competing on price or quality. Then one day, seeing an opportunity, a new grocer moves in with fresher fruit, a wider variety of products and most importantly lower prices. All the customers go over to that grocery store and the old grocers have the choice to improve or die. From the point-of-view of the previous grocery store owners the new grocery store is "greedy", but it ends up benefiting everyone else.
I think its more important for investment. If you have $1mil in cash and know it's losing value every day you have an incentive to invest it in some long-term profitable way. Hire more employees, buy some more trucks for your fleet, renovate your store, do some R&D to improve your product, etc. If it's the opposite you don't feel any urgency because your $1mil is gaining value as it sits in the bank.
I'm not an economist, so maybe someone more knowledgeable can weigh in. But my understanding is that deflation is worse. If you can just stick $10k under your mattress and expect it to be worth 10% more in a year you have no incentive to invest. Businesses will just hold their cash, banks won't have money to loan out and the sort of investments that provide new jobs, goods and services are a risky high-effort bet compared to just saving.
Most people make their ideas and opinions part of their identity. And so if it turns out they are wrong about something then a part of them has died, or at least severely injured and needs to be healed. A few people are not like this and their ideas are more like a collection of trading cards they keep in their pocket. They think they have a pretty good collection but are not opposed to throwing out one for another if they find something more valuable.
The latter types are the only ones who you can have honest intellectual debates with.
Yeah. Opus is nice for tasks that require significant planning and considering broader effects on other parts of the code. But it likes to go off the rails and do too much. Often it gives good-sounding ideas but it has a tendency to distract me by giving me a huge to-do list.
Yeah. Mine really likes to read excess code. I'll ask it questions like "If I move all these three ETL jobs into a subfolder will it break anything?" It'll start with giving me the simple answer but then continue on to consider another question and realize it requires reading my entire other repo that handles all of my cloud's infrastructure. And it'll proceed to read through tens of thousands of lines of terraform.
It's lack of trust
Yeah lol. That definitely does not sound like philosophy. Giving a "why" you want to implement a feature and make particular changes will help the AI stay on track much better than if it is driving blind. It can't make choices without understanding what the desired outcome is.
The fact that dozens of companies seem to have been able to produce great models, and the mid-tier ones are only months behind the frontier models tells me that just having a great model is not much of a moat. It may actually be having the physical stuff that makes all of this possible is the bigger revenue generator.
Saying it's like Pepsi criticizing Coca Cola is probably a better analogy.
if you're testing your water daily for tritium you'll catch it, but how often does that happen?
A good geiger counter can be bought on Amazon for $30. Something is either radioactive or it's not. You don't need to have a sophisticated understanding of the chemistry to test if dangerous radiation is present or not, nor do you need sophisticated equipment. My point is that being next to a radiation hazard should not cause the same sort of anxiety-of-the-unknown that living next to, for example, a chemical plant that may produce a menagerie of difficult-to-detect carcinogens.
From what I understand the waste problem is tremendously overblown. Move it to some storage facility somewhere, that's fine. Just keep it on site, that's fine too. A typical gigawatt reactor produces about 20 tons of waste annually, which sounds like a lot but remember this stuff is quite dense, so it would actually take 4 years to fill up a standard shipping container.
The storage units for this stuff is incredibly robust and safe. Radioactive stuff is also incredibly easy to detect. No company or reactor could ever leak into the community in a covert way. People would know right away. IMO, this is much less scary than being next to a chemical plant.
This was a good read, thanks for sharing.
I feel like the best way to fight this, if I was the government, would be to stop enforcing laws preventing patients and even doctors or hospitals from importing the overpriced pharmaceuticals from overseas generics manufacturers. Like the article points out, each pill costs no more than $0.25 to make. The goal wouldn't necessarily be to get everyone using overseas suppliers but to put downward pressure on the price gougers in the country.
There are a lot of government jobs I know of that are absolutely air-gapped. Your computer has basically no internet access, everything is stored on-prem. Hedge funds also tend to be extremely locked down, from what I saw when I interviewed. With certain data sets either having strict encryption-in-transit or a being stored in a quirky on-prem service. I can't imagine they're going to be dumping their data into Claude, etc.
As to why Windows is so dominant, I'm as clueless as you.
Yeah, on the H1B there seems to be a separate issue. Posting a job listing that is intentionally hidden or otherwise not going to get bites. That way a company can say that they couldn't find a candidate and have to hire someone in need of a visa. From what I understand they do this to undermine wages as well as having a ̶e̶m̶p̶l̶o̶y̶e̶e hostage who won't complain about poor conditions because they don't want to lose their visa.
Yeah. Going to be a lot harder to do this when other people in the company bring up that it is illegal.
I'm just trying to get it to help me learn physics. If my topic of interest shifts to mid-20th century European history I'll keep what you said in mind.
I think that's overstated, but the loss of trust companies have with the big AI players is pretty serious. Not a big deal if your app is for sharing cat videos, but if you're medical or wealth management or a government contractor or the like enterprise clients really like to see good data security policies.
Nice! Do you do anything with that compute when you're not actively using it? Is the crypto-mining hobby still worth it? I've also wondered if such expensive hardware can be rented back out to offset cost. Looks like these cards are going for as much as $4k nowadays.
And I can always see its full thoughts, don't have to worry about where my data is getting sent, and know it can't get secretly nerfed.
For this reason I wonder if local models are a potential business opportunity. Provide the service to engineering teams to give them a pre-built and setup GPU rig they can run in a closet. No need to worry about all the things you mentioned and clients can rest-assured their data isn't disappearing into a sketchy data center. There might be regulatory reasons that make on-prem setups appealing as well.
Very curious what hardware you're running this on!
I'm curious when coding-heavy companies will start running their own on-prem AI clusters. Has anyone had the idea to sell something like 4 GPU machine an engineering team could throw in a closet somewhere and run whatever they want on it? I imagine this won't appeal to everybody but with the trust issues the hyperscalers have developed hoovering up people's data and using it to train their models, I imagine some will find value in a machine and model they have transparent control over including the option to walk over and unplug the thing.