Maybe they've changed it in the past hour, but as I write this comment the 25$ plan is called "Ultimate" and promises unlimited search, but not unlimited assistant.
I agree about the need for appropriate wording and advertising, but other than that, the new limits seem entirely reasonable and in line with what other aggregators like Abacus and Poe are doing. The paid plans of the major AI labs themselves always have usage limits too. It simply can't work any other way if you include costly models in the mix.
"""
Aluminium nitride (AlN) is a solid nitride of aluminium. It has a high thermal conductivity of up to 321 W/(m·K)[5] and is an electrical insulator. Its wurtzite phase (w-AlN) has a band gap of ~6 eV at room temperature and has a potential application in optoelectronics operating at deep ultraviolet frequencies.
...
Manufacture
AlN is synthesized by the carbothermal reduction of aluminium oxide in the presence of gaseous nitrogen or ammonia or by direct nitridation of aluminium.[22] The use of sintering aids, such as Y2O3 or CaO, and hot pressing is required to produce a dense technical-grade material.[citation needed]
Applications
Epitaxially grown thin film crystalline aluminium nitride is used for surface acoustic wave sensors (SAWs) deposited on silicon wafers because of AlN's piezoelectric properties. Recent advancements in material science have permitted the deposition of piezoelectric AlN films on polymeric substrates, thus enabling the development of flexible SAW devices.[23] One application is an RF filter, widely used in mobile phones,[24] which is called a thin-film bulk acoustic resonator (FBAR). This is a MEMS device that uses aluminium nitride sandwiched between two metal layers.[25]
"""
Speculation: it's present use suggests that at commercially viable quantities it might be challenging to use as a thermal interface compound. I've also never previously considered the capacitive properties of packaging components and realize of course that's required. Use of Al O as a heat conductor is so far outside of my expertise...
Could a materials expert elaborate how viable / expensive this compound is for the rest of us?
I'm not much of an expert, but maybe this can be useful: AlN is a somewhat widely used insulating substrate that is chosen where sapphire is insufficient (~40 W/mK), but BeO (~300 W/mK) is too expensive or toxic. The intrinsic conductivity of single-crystal AlN is very high (~320 W/mK), but the material is extremely difficult to grow into large single crystals, so sintered substrates are used instead. This reduces thermal conductivity to 170-230 W/mK depending on grade. Can't comment on pricing though.
> The "real" point of an argument is not to persuade the other side (though that is what you aspire to nonetheless) but to exchange views.
Maybe this is just a matter of definitions, but for me the point of an argument is to convince or be convinced. When two incompatible views exist on a subject, at least one of them must be wrong. Some topics of conversation allow for diverging views or values, but then we are just talking or sharing experiences, not arguing.
That said, it is my experience as well that actually changing someone's (or my own) mind on an important issue is unlikely. Especially on complex topics with partial and uncertain information, like political issues, our life experience and cumulative knowledge significantly influences our selection of sources and interpretation of the facts, so converging on a common point of view may require the exchange of a prohibitive amount of information, even among rational arguers.
Productive argument usually occurs in a sort of semi-echo chamber, with people who mostly agree with us on the context, and are only arguing about the top layer, so to say. But when trying to argue about the deep stuff, we are mostly just "exchanging views", in the end.
> When two incompatible views exist on a subject, at least one of them must be wrong.
Rarely can any significant argument be boiled down to something so simple that this is the case. What if there are two incompatible views on the course of action that should be taken to lead to some desired outcome? You really can't just say that one of them must be wrong. There is a whole web of tradeoffs, assumptions, and odds to consider - you can't simply determine "right" and "wrong".
> When two incompatible views exist on a subject, at least one of them must be wrong
This isn't strictly correct if the source of incompatibility is differing assumptions / axioms. Both views can be correct in their own context and incorrect in the other context.
You underestimate how insane the power requirements of a high-g launch are. Placing one ton into orbit requires a rocket with close to 1GW peak power output (this is only a very rough figure that depends on mass ratio, thrust and exhaust velocity). Delivering this much power over 20 miles, even at 1MV, would take about a ton of aluminium wire (again, very roughly). The weight grows quadratically from there -- delivering this much power 200 mi downrange to the upper stage would take 100 tons, even setting aside all the other technical challenges. And this is where we actually want this power delivered -- we don't need to increase the ISP of the first stage, as it is mostly wasted in the exhaust anyway. An ISP above the capability of chemical propulsion is beneficial in the upper stage only.
A couple notes. Running your numbers through Ai it looks like it would weigh about one ton per km of wire. And possibly much less if you’re ok with more power loss and high wire heating. Since this is a short term use that might be fine.
But on top of that the extreme amount of power is only needed because we’re trying to get up to speed as fast as possible to minimize gravity loss.
If we’re not carrying our fuel for the first stage it’s conceivable we only need enough power to hover the unit and gradually get it up to top speed.
I know it’s still a far out idea but might be workable from first principles.
Jonathan Blow's "Preventing the collapse of civilization" [1] makes a similar point. It is easy to assume that, if we can build EUV machines and space telescopes, then processing stainless steel and manufacturing PCBs is baby stuff, and is just waiting for the proper incentives to spring up again. Unfortunately that is not the case -- reality has a surprising amount of detail [2] and even medium-level technology takes know-how and skilled workers to execute properly. Both can be recovered and scaled back up if the will is there. And time -- ten or twenty years of persistent and intelligent effort should be plenty to MAGA :)
> But the important question is - is it worth it? Should we be doing something more valuable instead?
It's hard to quantify. E.g. the CHIPS act is a strategic thing in case TSMC is disrupted for some reason. How valuable is insurance? How much useful work (and skill) do you ship overseas in exchange for promissory notes[0]?
> E.g. the CHIPS act is a strategic thing in case TSMC is disrupted for some reason. How valuable is insurance?
What do you think the odds are that the CHIPS act will make the US capable of building TSMC-level chips?
Calculating the value of an insurance isn't complex. First, you calculate (1) the cost of the insurance, then (2) a rough idea of the probability that you'll need it, and (3) the cost of the insured-against-event occurring.
The problem here is that no one in their right mind believes the CHIPS act will make the US capable of producing chips on par with those of TSMC, so it's not actually an insurance. More like a political show — see, we doing something!
People seem to want jobs with the macho kudos of manual labour, but with the physical comfort and salaries of email jobs, and I have some very bad news about that combination.
Those people need to watch a few episodes of Mike Rowe's "Dirty Jobs".
Also people need to stop saying "unskilled labor".
There is no such thing as labor without skills,
outside a category in an archaic way of justifying low wages.
This is a pet peeve of mine: yes there are unskilled jobs. Lots of them. The term is maybe slightly misleading, but there absolutely is a class of jobs that any able-bodied person could perform given at most a few hours or a few days of training, and they are qualitatively distinct from jobs that require education, specialized training, and/or months or years of experience to be considered proficient and productive in them.
That doesn’t mean people who work jobs in the former category deserve ridicule or disrespect. But the distinction is important because finding workers to fill an unskilled role is just a matter of finding said able-bodied person, while for the latter you need some kind of system of education, training and/or apprenticeship (either explicitly or effectively) to be set up and functioning to even have an industry that depends on those jobs.
Not everything is some silly game of political fighting through language. Some things we actually need terms distinguishing “this” from “that” so we can have real world conversations about them.
Mike Rowe is a shitty human being who delights and is gleeful about the idea of those folks in these same “dirty jobs” being paid less, forced to work more and harder with less safety equipment, and with less respect.
He hates the idea of people getting ahead in life with anything but the most extreme back breaking labor. That’s why he’s hardcore MAGA and makes such a big deal about trying to shit on folks who do desk jobs.
Fuck him, fuck his show, fuck the “good parts” where he tries to show you that being a garbage man is hard. If it’s really that skilled, the market will pay it as such.
I think it is pretty useful to be able to distinguish between jobs that don't require much education/training, and jobs that do. "Unskilled" and "skilled" are how we do that. Do you have alternative words you'd use?
IMHO, with the Big Tech boom winding down, what is more valuable for us to do? Manufacturing could prepare us for the next wave, whatever that might be.
> IMHO, with the Big Tech boom winding down, what is more valuable for us to do?
Tech isn't winding down; tech, as the sector that draws the most investment based on long-term development, had the biggest response to tight monetary policy designed to slow the entire economy down, but that response demonstrates that tech is where most of the marginal dollar goes.
> Manufacturing could prepare us for the next wave, whatever that might be.
Trying to work our way down the raw materials -> manufacturing -> finance/services ladder that countries usually try to work their way up for maximum prosperity in globalized trade isn't going to prepare us for anything other than lasting economic decline. And why would “manufacturing”—which you can't build generically, but only by specific, usually impossible to reallocate to a different use that isn't closely similar without sacrificing most of the value, major capital investments in particular subareas of manufacturing, prepare us for anything else even ignoring that we’d have to regress to do it?
Just because we ended the era of cheap money to try and stop runaway inflation doesn't mean the tech boom is winding down.
Look at everything that's happening with gene editing, in physics, with the jwst, with LLMs and robotics and computer vision, with alt energy sources, batteries, in material sciences, etc.
I mean this is such a myopic take. We are in just now in an era where people are now capable of finding needles in needlestacks.
You are confusing easily manipulated economic vibes that feel bad right now with the rapid approach of a complete overhaul of the human experience.
The U.S. has basically supported the strip mining of our economy by value sucking predatory investment firms. There is a reason why China have more robotics per capita in their factories than we do and it has to do with a complete failure in strategic thinking, long term planning and ultimately a hatred for our youth.
> gene editing, in physics, with the jwst, with LLMs and robotics and computer vision, with alt energy sources, batteries, in material sciences, etc.
These are tidal waves compared to the tech boom tsunami we experienced in the last 25+ years: enabling rapid communication of every human on the planet and democratizing access (anyone can create a app/website/etc to enable other people to communicate/make money/etc).
> where people are now capable of finding needles in needlestacks
Yes, exactly. all that is left is going after hard problems that impact the long tail.
I've seen this brought up with board games that are now primarily made in China, because injection molding is cheaper there especially for small quantities. The US could make the board game minis, but everyone who is capable of it in the US is producing high value high quality aerospace, industrial, medical parts. It's a waste of their time to produce small runs of toy parts.
mold making is also pretty complicated -- anything in the 1,000-1M parts produced will _probably_ be an aluminum mold (cheaper than steel) but they're still heavy and large to keep around.
I haven't met any injection molding shops in the US that do a huge amount of specialty parts like toys. The industry tries to get as many medical device jobs as possible.
I've thought about this and love board games. I don't want cheap plastic anymore. I want a reusable modular gaming system that let's me use more imagination.
Finishes are getting much better, especially with the high resolution resin based printers. But they are still slow and labor intensive compared to a "real" factory.
That's a crazy statement. It is clearly not true that every single person in the US capable of making board games now or in the future is instead already making high-grade aerospace and medical components.
Depends -- do you want the US to become a vassal state of China? That's the trajectory we were on. China is going to catch up rapidly on technology, AI, and services, and before a few months ago the US was going to continue falling behind in every other conceivable area.
That’s a hilarious thing to say considering our behavior towards trade lately. We’ve burned bridges with our closest trading partners and made everyone else uncomfortable to trade with us because they don’t know what the eventual tariff rate will be, or if it will change tomorrow. We’re retreating from the world stage, and guess who’s sitting there ready to take the reins. It’s genuinely the opposite of what you seem to want.
Depends on how evaluate what is valuable. E.g. here in europe a lot of people think subsidising local agriculture is not valuable and we should just import cheaper food. On the other hand, a lot of people agree that food security is kinda valuable by itself. And want similar security in more fields. In that sense yes, doing „low tech“ is valuable in the long run.
I've been thinking lately that we don't properly account for things like security. I've also been thinking lately that a lot of people have terrible ethics and are more than happy to engage in nepotism and or fraud. Don't know what to do about it personally, I just try to keep my needs small and be happy with what I've got while trying to prepare my own children to have some level of a good life.
More like common man does not think long term (and I'd say rightfully so). While democratic regime embraces populist hedonistic solutions.
Who cares about defense capabilities 10 or 50 years down the line? Lots of people in West had a good run outsourcing everything. But once there's nothing else to outsource and IP to sell... It's not gonna be pretty.
Next generations in West will have to work very hard to recover from this mess.
Have you ever heard any concrete strategies and plans regarding food security?
Wouldn't there be policies about how many calories should be produced in what form, how long can it be stored, what would a local ramp up look like if there was a global catastrophe?
What percentage of agriculture is really relevant to food security?
Those are just empty words so farmers can get their subsidies and go on to produce more industrial rapeseed oil.
As long as you have whole supply chain locally, you don't need to store too much.
The problem with agriculture is you can't really „ramp up“ it on a whim. That's why you need to keep it going and you can't just kick start your food production when outside suppliers start to blackmail you.
> In that sense yes, doing „low tech“ is valuable in the long run.
Sure. But how much tax money do you want to throw at entire industries to hide the basic fact that wages are lower elsewhere? Where do you want to take the labor away from? And where do you draw the essential/wasted subsidies boundary line?
Because in my view, Trump tariffs just ignore those very basic questions and don't even attempt to answer them.
It's perfectly reasonable IMO to throw 20 billion a year to agriculture, because that is a very essential sector. But doing the same for the textile industry? Ore/Oil refining? Steelworks? Chemical plants?
I don't wanna subsidies 20 non-essential industries just so that some former fast-food worker can assemble overpriced shoes inside the US (and labor demand from all those industries would drive up wages/costs in the fast-food sector, too, thanks to the Baumol effect).
I'm not against nurturing some important local industries, but Trump tariffs are a complete failure at achieving that IMO.
Don't want to make hypothetical shoes? Fine. One day soldiers may end up marching barefooted and loosing a battle though.
IMO the global economy eventually self-levels. Either you go up the chain so far that you eventually go off the rails by being unable to make basic stuff. And eventually being eaten by more hungry people with the basic skills. Or you keep yourself down by forcing yourself to not loose basic skills. Former gives you a short moment of glory with a high price for future generations. Later forces people to be more ascetic if that's the right word.
You misunderstand me. The US is making shoes-- just not as many as it imports from Vietnam or China. In fact enough shoes get made locally to export about 1$ billion worth of them (while ~$20 billion are spent on imports).
But I don't see the point in throwing billions of dollars from taxes at this industry just to make all those shoes here-- that is stupid (because the jobs that would create are not gonna be very desirable, they are gonna drive up costs all over by competing for labor, and that kind of protectionism is gonna invite retaliation).
The situation is very similar for a lot of industries.
I also think it is extremely unhealthy to baby an industry long-term by isolating it from competition like this.
I'd be totally on board if there was like 20% unemployment in the US, and this was a short term plan to give those people work/income.
But that's not it. This is in my view really bad policy driven by emotional arguments, and actual numbers, expected outcomes and historical precedent (for "I know better than market economies what ought to be produced") all heavily weight against this.
I'm very confident right now that the whole "20%ish tariffs for everyone to balance trade deficit with everyone" approach is gonna be walked back or lead to abysmal outcomes, and people should have realized that from the start.
> In fact enough shoes get made locally to export about 1$ billion worth of them
We have far more shoes than we need.
> the jobs that would create are not gonna be very desirable, they are gonna drive up costs all over
Only because our government is run by billionaires. Elect politicians that care about the median American and this problem can be resolved quickly.
> I also think it is extremely unhealthy to baby an industry long-term by isolating it from competition like this.
This “babying” you mention results in decent working conditions and guaranteed jobs for Americans. It’s a trade off I think is worth it, as your proposal disproportionately benefits the 1%.
> I know better than market economies what ought to be produced
Have you looked at the astronomical surplus of useless goods we have here? Those come at the cost of labor that could be put towards jobs that benefit all Americans (building more homes, cheaper childcare, cheaper food, etc). Again you’re arguing for a status quo that is designed to grow the wealth gap and make billionaires richer. Essentially trickle down economics.
I think large scale modeling and allocation for "more valuable" has been overly narrow - insufficiently diversified for uncertainty/unknowns, and subtly incorrect for western nations for decades now
It is if war is in the future. And I’m not saying this as hyperbole but based on statements made by NATO secretary general (both Rutte, previously Stoltenberg and former General Bauer) about Russia’s military production outproducing NATO, or Finish President Stubb speaking on the powers of the world shifting and the need to ramp production which were echo’d recently by Macron, or the Arctic region soon to become a contested region with China and Russia attempting to stake their influence in the area which is obviously at conflict with the personal interests of the other countries in the region.
It seems obvious to me that the world is a bit hotter than before 2022, with the likelihood of some conflict between powers of the world coming to pass being greater.
If production of raw materials to usable materials is all contained within countries that are deemed to be unfriendly by the one lacking this production capability, it’s a clearly in their vested interested to not be in that situation.
Only problem is there is a seemingly idiotic US administration attempting to address these deficiencies, unless there’s some weird 4D chess play going on, but I’m not convinced it’s that.
The EUV light sources are all made in San Diego. Currently, there is no single country that can make an 3600D or equivalent machine. Which shouldn't be surprising given the complexity.
The article specifically mentions that OpenAI is losing money on all plans, including Plus and Pro. The 500 million number is strongly inflated by those who only use the service on and off.
Equity compensation is an essential part of modern corporate incentive structure. In particular, it incentivizes prospective employees to accept lower compensation, by making it appear larger on paper.
I've always evaluated it at 0, and that's all I got from the equity I got in my whole career. If I didn't think the salary was enough I wouldn't have accepted.
It would probably have been more correct to say "requires fewer allocations in some cases". As you point out, in terms of layout, the old generic version is just as intrusive as the new version, and requires just as many allocations (one). However, the new version gives extra flexibility for you to move a long-lived object in and out of lists without copying or allocating, at the cost of having the pointer field baked into it permanently, rather than on demand.
I think the reasoning behind this change is that (from Zig's perspective), if you are using linked lists you are probably doing something wrong, unless your application requires the above-mentioned kind of juggling, which favors explicitly intrusive LLs. In addition, this change de-generifys the lists's methods, like `prepend`, which reduces code size a little.
> However, the new version gives extra flexibility for you to move a long-lived object in and out of lists without copying or allocating
You could also do this with the entire node in the case of a generic implementation though, the API just needs to expose the node struct to you and allow you to detach it; but the same is true for this new implementation as well.
In terms of memory, a detached node that embeds the payload struct isn't different from an object with an embedded detached node.
What changes is that now, if you have an object class that you don't want to (or can't) extend to include a list node, you have to wrap it in a container struct that, again, looks the same in memory but now has a node and your object as its members. I'm not sure if this is really much of an improvement at the end of the day.
Also, correct me if I'm wrong (I don't do zig), but shouldn't it be possible to create a list of void elements from the generic implementation and embed its node type inside your object, then proceed from there as if it were the new implementation?
Yeah... with bitcasts and some creativity (and some boilerplate) both versions are ultimately equivalent, or nearly so. But the new one pushes you towards intrusive-data-structure thinking and away from container-of thinking.
This, by the way, is a typical design consideration in Zig -- using "friction" to steer programmers away from doing the Wrong Thing (according to Andrew). In addition, Zig is really more of a fancy macro-assembler for outputting optimal code, and less a pragmatic general-purpose language, and makes design decisions accordingly. Taking both together, the linked-list change sort of makes sense, even though personally, I would have just added a separate intrusive list data structure.
This looks pretty bad. I was naive enough to expect that the pro-Elon faction would cancel useless garbage like Artemis and Starliner, but they instead go after the actual science programs?
I agree about the need for appropriate wording and advertising, but other than that, the new limits seem entirely reasonable and in line with what other aggregators like Abacus and Poe are doing. The paid plans of the major AI labs themselves always have usage limits too. It simply can't work any other way if you include costly models in the mix.