When :extend-via-metadata is true, values can extend protocols by adding metadata where keys are fully-qualified protocol function symbols and values are function implementations. Protocol implementations are checked first for direct definitions (defrecord, deftype, reify), then metadata definitions, then external extensions (extend, extend-type, extend-protocol).
I asked it to make a drawing of the US with every state numbered from biggest to smallest with 1 being the largest.
Maine was #89 (That is not a typo.) and Oregon was #1.
OpenAI as a company simply cannot exist without a constant influx of investor money. Burning money on every request is not a viable business model. Companies built on OpenAI/Anthropic are similarly deeply unprofitable businesses.
OpenAI needs to convert to a for-profit to get any more of the funding that Softbank promised (that its also unclear how Softbank itself would raise) or to get significant cash from anyone else. Microsoft can block this and probably will.
It all reminds me of that Paddy's Dollars bit from it's always sunny.
"We have no money and no inventory... there's still something we can do... that's still a business somehow..."
I tried a fairly basic Pokemon Go question - which pokemon are resistant to ghost attacks, and it got it wrong - said normal types are immune. Which is wrong. ASI is not quite with us yet.
burning money worked for Uber. As long as they can IPO or get cheap debt from governm friends any valuation can work. Uber lost double digit billions as an app with no edge or anythin. It always made no sense beyond 1 billion
Uber's whole schtick is being what was an already profitable business model (Taxis) with lower overhead/easier access.
That money they burned was on customer acquisition, building infrastructure, etc. The unit economics of paying to be driven to the airport or Benihanas was always net positive.
They weren't losing money on every customer, even paying ones. There just isn't a business model here.
I wouldn't say they had no edge. They had a huge advantage over traditional taxi companies. You can argue that a local Uber-like app could be easily implemented, that's where the investors came in to flood the markets and ensure other couldn't compete easily.
The situation is in no way similar to OpenAI's. OpenAI truly has no edge over Anthropic and others. AGI is not magically emerging from LLM's and they don't seem to have an alternative (nobody does but they promised it and got the big bucks so now it's their problem).
the 1.5 mil bonus to tech staff announcement prior to chatgpt 5 release makes even more sense now. They knew it would be difficult to manage public expectations and wanted to counter the short-term (in the best case) drop in morale in the company.
I think that 1.5m bonus is likely stocks with 500B valuation. There are other rumors they want outsiders to be allowed to buy stocks with 500B valuation.
The way they do this in Europe is that an enterpreneur buys a fleet of cars and then gets a visa for a number of folks from Bangladesh and other areas who don not own any of these cars and ride them in turns (they also sleep like 10 in one appartment but that's a different story). The owner gets the money and distributes them to the actual drivers. Uber says they are innocent as they are not in an emplyer-employee position with any of these drivers.
This model worked for the fleet owners so far because the Saudi gave enough money so that both (1) the customers were happy, (2) the cash from the ride could be divided between owners and drivers in a way these drivers complained only to a certain extent.
But the last two years (the only profitable ones) are much worse, both for the drivers and fleet owners. There is still sunk cost in there, but once the cars get old enough they will need to think well whether to buy/lease the next batches.
Uber raised something like $50 billion in debt and equity before it went public, but after 15 years of losing money, it has finally started making profits… just in time for Waymo to arrive and eat its lunch. Of course, Uber could themselves get into the self-driving game, but their entire profit story to investors relies on pushing costs away from them onto drivers; it vanishes entirely if they have to maintain their own fleet.
Uber is profitable on a cash basis, but if you’re a public investor, you got fleeced by the early-stage venture money and debtholders. I don’t think it will ever pay back what it raised.
"...while Uber has achieved profitability, some analyses suggest that a substantial portion of these profits may come from an increased revenue share at the expense of drivers' earnings".
So let's imagine is 2040 and OpenAI is finally profitable. Now, Uber did this by increasing prices, firing some staff and paying smaller wages to drivers. And all this while having near-monopoly in certain areas. What realistic measures would they need to take in order to compete with, say, Google? Because I just wish them good luck with that.
I had it create a map "in the style of a history textbook." It came up with something that looks worse than I imagined: https://pasteboard.co/3zGy5ti4hHuT.jpg
Isn’t it old news that the full for-profit is not happening and they renegotiated the terms that would make the current proposed PBC a solution as it meets the economic terms?
I have no idea if OpenAI succeeds or not but I find arguments like yours difficult to understand. Most businesses are not using these systems to draw a map. Maybe the release of 5 is lackluster but it does not change that there is some value in these tools today and ignoring R&D (which is definitely a huge cost) they run at a profit.
> ignoring R&D (which is definitely a huge cost) they run at a profit.
how can you say such a hand wavy comment with a straight face? you can't just ignore a huge cost for a company and suddenly they are profitable. that's Enron level moronic. without constant R&D, the company gets beat by competitors that continue R&D. the product is not "good enough" to not continue improving.
if i ignored my major costs in my finances, i could retire, but i can't go to the grocery store and walk out with a basket of food while telling them that i'm ignoring this major cost in my life.
I don’t know why so many take these discussion with such a high emotional level. Has the ability to constructively discuss a topic been lost? I know you usually respond with high emotion and brash but at least try to be constructive.
It’s a valid point and that’s the biggest question when it comes to the medium to long term business plan. Those R&D costs are an important part of it. My point is that since runtime is profitable there is a lot more runway to figure out how to tweak R&D spend in such a way that it becomes a viable business for the long term.
There are a lot of questions that they need to answer to get to pure profitability but they are also the fastest growing company on a MAU number in history with a product that you can see has a chance at become profitable from all sides. They may fail or become sidelined but the hyperbole and lack of critical discussion here is disappointing.
I like how when your illogical notion is challenged, you respond by saying the challenger is being emotional.
There is no point in saying that an AI company can just ignore its R&D. There is no company without the R&D. Because of that, any conversation pretending it doesn't exist is pointless. There is no constructive conversation with that as the premise.
You’re arguing against a point I’m not making. I’m not saying R&D isn’t necessary or that it “doesn’t exist”, I’m saying that operationally, the service itself runs at a profit before accounting for R&D. That matters because it means they have a viable revenue engine that could, in theory, fund a sustainable R&D budget if they adjusted spend.
That’s a very different conversation than “pretend R&D doesn’t matter.” No one is suggesting they stop building; the question is whether they can align the burn rate with the revenue base over time. Companies make those tradeoffs constantly when maturing from heavy investment to profitability.
And yes, you are being emotional, not because I disagree with you, but because your language is inflammatory and brutish. It’s hard to have a constructive discussion when every response is dialed to 11. Misframing the premise as “ignoring a huge cost” isn’t debate, it’s a straw man, and it sidesteps the real question of whether the underlying business model works once R&D is right-sized.
Would love to have a real critical discussion on why you disagree but please leave the bad language out of it. It’s boring and I know it’s your typical route in these types of discussions but at least have a valid retort.
Correct, past R&D spend is already sunk and can’t be undone. But that’s why it’s useful to separate sunk costs from future operating costs when evaluating viability.
The relevant question is whether the ongoing revenue from the existing product is strong enough to support a sustainable level of R&D going forward. If your runtime margins are healthy, you have options: scale back R&D burn, focus on incremental improvements, or use the profits to fund more ambitious projects.
The entire US stock market is propped up by big tech companies spending massively on Data Centers and GPUs for AI. OpenAI is valued higher than Netflix.
A company that can pull in single digit billions in revenue for hundreds of billions in expenses just doesn't make sense.
> Most businesses are not using these systems t̶o̶ ̶d̶r̶a̶w̶ ̶a̶ ̶m̶a̶p̶.̶
FTFY
And no - while it might be obvious from the outside in that it probably won't happen, the continued existence of the business is still predicated on conversion to a for-profit. They don't just need the amount of money they've already "raised", they need too keep getting more money forever.
FTFY? Cute, but you’re arguing against a strawman. My point wasn’t that companies are using GPT to draw maps, it’s that dismissing the tech based on one goofy output ignores the far more common, revenue-generating use cases already in production.
As for “single-digit billions in revenue vs. hundreds of billions in expenses,” that’s just bad math. You’re conflating the total AI capex from hyperscalers with OpenAI’s own P&L. Yes, training is capital-intensive, but the marginal cost to serve (especially at scale) is much lower, and plenty of deployments are already profitable on an operating basis when you strip out R&D burn.
The funding structure question is fair, the for-profit conversion path matters but pretending the whole business is propped up solely by infinite investor charity is just wrong.
Microsoft AI Revenue In 2025: $13 billion, with $10 billion from OpenAI, sold "at a heavily discounted rate that essentially only covers costs for operating the servers."
So I looked up DBT - it seems to be a company where the first thing I saw on their website was "Gen AI."
That doesn't inspire confidence in the longevity of their offering - but i'm also unclear exactly what it is they are offering. Can you give the audience the pitch?
Sure, DBT (Data Build Tool) has been around for a while. It's a way to manage SQL (and also Python) data pipelines - the "T" in ELT. Home page for DBT Core is here https://docs.getdbt.com/ (there is also a hosted DBT cloud premium offering). There are some good books on Data Engineering with DBT - I particularly liked "Data Engineering with dbt: A practical guide to building a cloud-based, pragmatic, and dependable data platform with SQL" by Roberto Zagni.
So an intrinsic problem with flattening the database is that a lot of applications actually do want to work on a tree of information (that's what the json query part is all about)
Dropping a view can work, but not when the system is under load (found that out the hard way.) I'm not saying its the worst possible situation, just a greater burden than you would expect going into it.
Can you elaborate on `instead of` triggers? How do you make use of those?
Yeah - in another place this was shared a few people mentioned that. It is slated to be part of postgres 18, but as of right now to get the "good UUIDs" you need to homeroll something.
Thats the only reason I didn't mention it, seemed a bit of a rabbit hole.
This is a book intended to teach someone the Java language, from scratch.
You will find that the content makes heavy use of recently released and, for the moment, preview features. This is intentional as much of the topic ordering doesn't work without at least Java 21.
Right now I have several key areas where I could use some help:
* Writing Challenges. Most of the early sections have challenges students can do to test their understanding of the topics covered and for practice. I've shifted my focus away from these to make more progress on the main content of the book. Any assistance would be appreciated.
* Theming. A lot of the chapters are...bland. Purely technical. I find that when I have the imagination to "theme" the subjects they become higher quality and more engaging overall. See an anime you liked recently and think you can make the math chapters use the characters from it? Give it a shot!
* Fixing Mechanical Issues. I don't have an editor and I don't often proofread. If you find mechanical errors in my grammar or find issues with the way topics are ordered I would welcome fixes.
Notably I do not want to open the floodgates for contributions on the main chapter content just yet. This has the downside of slower progress but the upside of a more coherent result.
My primary goals with this are
* Get the ordering of topics right. By this I mean that every topic covered should have had its prerequisites covered in the topics previous. While "lesson 1: Inheritance" is clearly wrong in this regard, some things are more subtle.
* Be a template for other people. This is a book. Not everyone likes books, some like youtube videos, some like over priced udemy courses, some attend College, etc. Everyone has different learning paths. I hope this to be of use to anyone looking to make a more up to date Java curriculum and hope that the vague order of things (which I consider superior to the content produced with the Java of years' past) is carried through.
* Write as if the newest Java wasn't new. It's obvious when a book was written before Java 8 because it always has newer additions with "addendum: brand new stuff in Java 8." But the order language features were introduced is hardly a good order to teach them. You have to pretend that Java 23+ has always been the Java. Does it really make sense to show terrible C-style switch statements way before switch expressions?
* Write as if the words Object Oriented Programming, Functional Programming, etc. didn't exist. While I understand that these all have definitions and are useful concepts to know about, introducing them early seems to lead to either dogma, rejection of said dogma, or some mix thereof. None of them are actually needed to understand the mechanics of and motivation behind what we would call "object oriented" or "functional" techniques. They certainly don't work as justification for adding getters and setters to every class.
My immediate short term goal is to get this "ready to go" for when anonymous main classes is in a stable Java release. Thats the point at which we could start to:
* Have actual students go through it without also needing to explain the --enable-preview mechanism.
* Use the topic order to build other sorts of non-book resources like videos, curriculums, projects, etc.
* Convince actual teachers to change from "objects first" to something less insane.
I haven't integrated println or readln yet, but will do so eventually.
so like if the funding for AI disappears then somehow my requirement of multimodal search also disappears, and with it all the existing solutions, some NOT VC funded, like pgvector?
