Anthropic has a great product, but what's going on in the stock market is astonishing. Companies waiting to be valued at a trillion dollars before going public? (I'm writing this comment with the assumption that they will go public soon and the valuation will be higher than this $965 billion dollar private valuation) The stock market used to be a place for companies to raise money from investors. But that isn't what it is anymore, it's a dumping ground. Venture capitalists & private investors are sucking all of the possible growth and future upside from these companies and then dumping them on retail investors when there's nothing left. There is no growth or upside left by the time these companies go public. If you invest in these IPOs you are buying the absolute peak with all potential future profits baked into the price, with nowhere left to go but down.
I shared the author's frustration when figuring out how to ship such binaries to end users so I wrote a guide [0] detailing exactly how to do it. Apple's documentation is surprisingly poor and I couldn't find any blog posts so I ended up reverse engineering what works via trial and error as well as popular OSS projects on GitHub.
Oh it gets worse than that, the money which caused all of this by OpenAI was taken from Japanese banks at cheap interest rates (by softbank for the stargate project), and the Japanese Banks are able to do it because of Japanese people/Japanese companies and also the collateral are stocks which are inflated by the value of people who invest their hard earned money into the markets
So in a way they are using real hard earned money to fund all of this, they are using your money to basically attack you behind your backs.
5000 years of civilisation and a sizeable chunk of the population of first-world countries still doesn't understand why law and courts are better than "just kill people who are le bad".
Let's not even talk about the killing of shipwrecked people who are hors de combat according to the Geneva Convention. Even the Nazis during unrestricted submarine warfare had the decency of not murdering shipwrecked crews and passengers. We are truly in a wretched time when nothing matters anymore.
I get all those same ads on YouTube. It’s easily >90% complete scam products, and most of the non-scam products are at least breaking YouTube’s own “rules” regarding what content is allowed. Like if one of the reportable reasons is “addictive product” why do I see so many ads for nicotine products?
I tried to report an ad that had literal nudity/porn in it. The report page was acting funky so I pulled it up manually in safari where I wasn’t logged in. and it made be log in to report the ad because the ad was age restricted. why is it running as an advertisement at all then? Clearly it got flagged somewhere to be age restricted.
YouTube has approximately zero incentive to fix this. They are a complete monopoly and there isn’t any consequence to any of the blatantly illegal products that I see advertised on their platform. I’ve seen ads for drugs and firearms. The firearm ad was claiming it was “easy to sneak past security” but the highest consequence is that the ad account gets nuked and another immediately takes its place.
I'm going all in on my side project CodeBrew, a Java IDE for iPad. Currently working on OpenGL support for 3d graphics, as some schools requested the feature. Also I'm finally pitting some work into aquisition, which has turned out to be much more fun than I anticipated.
Go check it out, its free to try, with a one-time purchase full version:
>Also, in addition to planned obsolescence and repair hostility is Design for Manufacturing (DfM) that doesn't care about maintainability, safety, comfort, or durability, only lowest cost to shove things together on an assembly line.
Exactly. It's basically fight club math. Spend $10 on a click-fit connector that can't be disassembled but that a $60/hr (though they only see a fraction of that) UAW laborer can plug in in half the time can't easily short-insert that can be visually checked.
The fact that it costs $200 the 1/10000 times it fails under warranty doesn't matter with those numbers. And you don't even care about the 100/100 times it fails at 2-3x the warranty period.
Of course, you're burning credibility doing this. But credibility doesn't have an obvious mapping to a number and stonk go up, KPI go up, bonus get paid, nobody cares.
Let's assume you're not a FAANG, and you don't have a billion customers.
If you're gluing microservices together using distributed transactions (or durable event queues plus eventual consistency, or whatever), the odds are good that you've gone far down the wrong path.
For many applications, it's easiest to start with a modular monolith talking to a shared database, one that natively supports transactions. When this becomes too expensive to scale, the next step may be sharding your backend. (It depends on whether you have a system where users mostly live in their silos, or where everyone talks to everyone. If your users are siloed, you can shard at almost any scale.)
Microservices make sense when they're "natural". A video encoder makes a great microservice. So does a map tile generator.
Distributed systems are expensive and complicated, and they kill your team's development velocity. I've built several of them. Sometimes, they turned out to be serious mistakes.
As a rule of thumb:
1. Design for 10x your current scale, not 1000x. 10x your scale allows for 3 consecutive years of 100% growth before you need to rebuild. Designing for 1000x your scale usually means you're sacrificing development velocity to cosplay as a FAANG.
2. You will want transactions in places that you didn't expect.
3. If you need transactions between two microservices, strongly consider merging them and having them talk to the same database.
Sometimes you'll have no better choice than to use distributed transactions or durable event queues. They are inherent in some problems. But they should be treated as a giant flashing "danger" sign.
Is there a good general purpose solution where I can store a large read only database in s3 or something and do lookups directly on it?
Duckdb can open parquet files over http and query them but I found it to trigger a lot of small requests reading bunch of places from the files. I mean a lot.
I mostly need key / value lookups and could potentially store each key in a seperate object in s3 but for a couple hundred million objects.. It would be a lot more managable to have a single file and maybe a cacheable index.
It's not just price discrimination, but the process of applying and choosing a college is essentially an auction process designed to benefit colleges over students. US colleges all band together to participate in a single coordinated system (FAFSA, college ranking systems, the common app, Early decision rules, etc.) , acting almost like a cartel, so they get away with it. If you view the college application process through the lens of Mechanism Design you can see that colleges hold all the cards and so the entire system is designed to their benefit: https://en.wikipedia.org/wiki/Mechanism_design
What exactly is unfair about the process?
1. The colleges which most heavily implement price discrimination are the most desirable to attend and also tend to accept large portions of their student body through Early Decision applications that don't allow apply to most (not all) of the other most desirable colleges. This means that accepted ED applicants have no recourse beyond either accepting or rejecting the proposed "aid package".
2. Highly desirable colleges aim for high applicant "yield" for rankings/planning, which in aggregate makes it so most of their accepted students only get into one highly desirable college. Even outside of early decision applicants, this puts applicants in a bad bargaining position - they must choose between either paying more for the ~single highly desirable college they got into or less for a less desirable college.
3. You cannot generally bid between colleges even if you got into comparably desirable/expensive desirable colleges (you can't tell Dartmouth and Brown that each is proposing a cost of $40k/y and have them bid against each other). This is because they essentially operate as a cartel. This limits downward pressure on prices.
4. The application process operates in rounds with fixed dates, there aren't really do-overs within a given year, and waiting for the next year changes the process (you're either a transfer or gap year applicant). This puts a lot of pressure on applicants to accept the least-worst option and doesn't give them the ability to react to a bad outcome by eg applying to more places after the fact.
5. The acceptance criteria are opaque and in many cases subjective (eg your application essays). Applicants need to hedge their bets and deal with a lot of uncertainty. Any accepted offer from a highly desirable college then feels like a gift and not worth squandering/negotiating.
6. As you mention, colleges know exactly how well you'll be able to pay, and because they act as a cartel that disallows bidding wars or negotiation + all the other forced scarcity/time pressure I mentioned, they can essentially extract as much from applicants as they want, up to their sticker price.
It's worth mentioning that not all competitive colleges participate in the cartel to the same degree as the Ivy League. When I was applying to colleges, I remember MIT and Caltech had Early Application (not Early Decision) processes, didn't make applying in those rounds as beneficial as ED colleges, and didn't have as many athlete/legacy "backdoors" as the Ivy League.
I also remember that Duke and Vanderbilt offered full merit scholarships to some students who might've received no need-based aid, which I was fortunate enough to benefit from and am extremely grateful for, even if it was probably a self-serving policy to poach applicants away from the Ivy League/improve yield.
Hey, if you're looking for a real-world pragmatic and performant implementation of a theoretically-cool algorithm, my https://github.com/tpn/perfecthash project might fit the bill.
It's geared to generating perfect hash tables with the fastest possible lookup/index times (for 32-bit keys), for key sets in the <=100,000 range. (It scales well up to millions of keys, but the solving time takes a lot longer.)
Nice polish on the app. However, I built this exact thing for myself using PowerShell (you could use any scripting language) , my OpenAI API Key and Snippety (on macOS). It was trivial to do and the costs are next to nothing.
But for users who want something ready made and don’t mind subscriptions this looks nice.
