This matches my experience. Burned $2K to see how it will perform on frontend tasks and backend tasks.Frontend did a significantly better job than Opus on toy-scale wireframe projects by using gimmicks like fluid dynamics. Then when given medium to big tasks like multi-page web app where layouts and aesthetics must be decided by model itself, results by Fable and Opus scored indistinguishable score from human judges.Backend, gave tasks related to setting up a data flow that involves Postgres, R2

A few things. I replied to someone else above, but I feed lessons learned from my social ant farm agents back into more productive agents.Memory recall:Lots of systems out there to give agents memory. I've used a bunch and written a couple. Storing memories is easy, but getting an agent to recall them, no matter how much you mention it in your AGENT/CLAUDE.md is a bit of an uphill battle. I've even watched claude make useful project memories and never refer to them again.In my agent ant farm

What do you mean by succeed? To me, it's very specific. It is to live without debt or a mortgage in a safe place where I can afford medical insurance (which currently costs me $65,000/year). It also meant being able to spend as much time with them as my children required.I guess you could say it took me from ages 21-38. If you count study as a child, 12-38.However by the time I was 23 I was already super happy. I had a good job and could afford to take my girlfriend out a couple times a week. So

This is so exactly right and I've been saying it to whoever will put up with me...(and now am embarrassed I have no link to show for it. oh well, shame is good for writing. envy too!)Software production is now so easy that everything is a .emacs file (pronounced "dot emacs" btw): meaning, each individual has their own entirely personal, endlessly customizable software cocoon. As tptacek says in the OP, it's "easier to build your own solution than to install an existing one" - or to learn an exis

This is all way too much. If you see a duplicate idempotency key, skip the replay and always return 409. This becomes a client problem. Clients already need to help enforce idempotent contracts; "check for conflict response" is not an onerous imposition.I've built multiple ecommerce APIs with this approach and they work great. No heroic measures required. You can often satisfy this contract with a unique constraint; if not, a simple presence check in redis. No hashing or worrying about PII.My ra

The worst part is the sharp changes in the price being traded aren't achieved by magic but rather with guns & actual human suffering

A little update: I added privacy-focused optional shorter URLs to SDocs.You can read more about the implementation here: https://sdocs.dev/#sec=short-linksBriefly: https://sdocs.dev/s/{short id}#k={encryption key} └────┬───┘ └───────┬──────┘ │ │ sent to never leaves server your browser We encrypt your document client side. The encrypted document is sent to the ser

> if we had forced the peaches to be grown and canned (as many comments are suggesting) then that would be a different kind of waste as they'd sit in warehouses while the land, resources, and labor were used to produce something people weren't buying instead of being used to produce foods they were buying.Worse, the price would have to be lowered to bring up sales, which could put the other peach farmers into bankruptcy as well.

This question has multiple layers of thinking:1. People who can't read pick randomly.2. People who can read, but are too dumb to model or care about other people pick red.3. People with enough intelligence for basic cognitive empathy pick blue.4. People a little smarter and think through game theory overall pick red, and think they are smart for doing so.5. People smarter than #4 and capable of seeing the big picture realize they don't want to leave people who choose #1 and #3 dead, so they pick

The US government is not currently in a state of following the law or constitution. People get fired, and if authority was not there, a lawsuit 9-18 months later might rectify it, and in the meantime the fired employee has moved on. DOGE cuts were extreme, capricious, and the only rhyme or reason was to try to hyperpoliticize the science to meet what people were guessing that Trump would want. On the grant side, they cut grants in an explictly racist way, according to a Reagan-appointed Republic

> Premature optimization is the root of all evil.There are few principle of software engineering that I hate more than this one, though SOLID is close.It is important to understand that it is from a 1974 paper, computing was very different back then, and so was the idea of optimization. Back then, optimizing meant writing assembly code and counting cycles. It is still done today in very specific applications, but today, performance is mostly about architectural choices, and it has to be given co

This is what happens when you allow money to influence power without check.What can be done to curtail it? Ban corporate donations to political parties and PACs. Limit personal contributions. Implement campaign spending limits so parties can't spend hundreds of millions on an election if they somehow manage to get that much money.Other nations (e.g. Canada) do this. It's not perfect. Money is always looking for a way, and politicians are always looking for the kind of power that money buys.

Do not try to solve an unsolvable problem, you'll end up hurting real users quite a bit more than you might imagine. Imagine new enthusiastic users trying your platform getting hit with an AI label because of inevitable false positives.'Detecting AI' is not a problem that has real solutions, the only avenue is something supply side like synthid. But that harms users too, by introducing further barriers for indie users.

I think a lot of the reason for the war on taxes is the exorbitant privilege [1] of owning the world reserve currency. It lets America print as many dollars as it wants, and borrow in a currency it controls entirely. In a normal country this would result in severe inflation, but because America borrows and prints a currency that is necessary abroad to conduct international trade, it is able to "export" a large part of its inflation.In such a system, it is rational to cut taxes as much as possi

> Always judge an author by the length of their text.Flashbacks to a past employer where the CEO decided that brevity was a core company value and started rewarding people for short communications and scolding us for longer text.Over the next year a few charlatans moved up the ranks by spitting out half-baked thoughts and e-mails all the time, which looked like clarity and brevity on the surface. People were afraid to speak out or discuss nuance because it was too many words, and you didn't want

So, when I code review, I have a super simple Cursor command that "orients" me in the PR:* where does the change sit from a user perspective?* what are the bookends of the scope?* how big is the PR?* etc.Once I'm "in" and understand what it does, I pepper the AI with questions:* Why did the author do this?* I dont understand this?* This looks funky, can you have a look?* etc.The more questions I ask, the more the AI will (essentially) go "oh, I didn't think of that, in fact, looks like the issue

The horizontal control of venues is only one issue. A perhaps bigger issue is the vertical integration (if that's the right term) of first-party ticket sales and resale in one company. Ticketmaster has no real incentive to try to prevent resellers from buying up all the tickets on first sale, because it gets to charge fees on all the resales through its platform. The more times a ticket is resold, the better.I don't believe a court would ever mandate this, but I'd like to see tickets sold by

Why not just ban the transfer of tickets and allow refunds? You buy a ticket, you show your ID at the door. Early refunded tickets get resold online and late refunds are sold at the venue. All seats, including the best seats, go to actual fans instead of scalpers just hoping to make a profit while providing zero value. First choice in seats goes to the most passionate and attentive fans.

Very good move. In my experience, for system programming at least, GPT 5.4 xhigh is vastly superior to Claude Opus 4.6 max effort. I ran many brutal tests, including reconstructing for QEMU the SCSI controller (not longer accessible) of a SVSY UNIX of the early 90s used in a 386. Side by side, always re-mirroring the source trees each time one did a breakthrough in the implementation. Well, GPT 5.4 single handed did it all, while Opus continued to take wrong paths. The same for my Redis bug trac

I have been a parent since I was 15. Officially married and moved on our own at 19. Graduated from the university at 22. Struggled hard-core until about 30 when my career changed and finally kicked off. My wife became an at home mom for our now three kids. It was my 40s when I realized, "oh, others see me as the adult in the room." I joke and say, "i have always been in my 30s," but I do feel a change recently. Very much facing forest dweller stage already with my oldest getting married.What mak

Agreed. QA specialists are there to think about what the engineer didn't think about. Unless the engineer is incompetent or the organization is broken, the engineer has already written tests for everything they could think of, but they can't think of everything.More importantly, it is almost impossible for engineers to be as well incentivized to spend extra time exploring edge cases in something they already believe to work than to ship a feature on time.Like everything else though, its contextu

In case you're curious, when I ran that title/thumbnail AB test, the option "This picture broke my brain" did end up winning. I was a bit disappointed, because I didn't really _want_ it to win, but I did include it out of curiosity. Ultimately, I changed it to the other title, mostly because I like it better, and the margin was small.I was genuinely torn about how to title this, because one of my aims is that it stands to be enjoyed by people outside the usual online-math-viewing circles, especi

Its five years with no limitations, so when you are due to be released; Whats your password? Another five years... Its such a poorly worded law you could literally spend your life in prison for forgetting your password. And Its mostly used against peaceful protesters.

I can foresee a design flaw, which is that the cat will ignore all the specially designated areas and sit on your keyboard instead.

A few things I do: I'll point something out, and ask a question. So if I'm in a shop I'll saying something like "I've never seen that before - is it popular?"If I'm getting a coffee if the barista says "How are you?" Rather than just saying something non-committal I'll say "I'm good thanks, it looks like you're having a busy day/quiet day - has it been like this all day?" or I'll ask a question about the beans (if it's the sort of place they regularly rotate through different beans) or I'll ask

I don't know if I like Anthropic more, but I certainly like their competitors much less now.The new thing that I know about leading AI companies that aren't Anthropic (i.e. OpenAI, Google, Grok, etc) is that they knowingly support using their tools for domestic mass surveillance and in fully autonomous weapon systems.

Street View is such a missed opportunity. In 2007 it was visionary and essential to create the map data that allowed Google Maps to win. In 2026 it is a symbol of Google's stagnation. Essentially zero improvement in user experience for more than a decade, in a time of incredible advancements in computer vision.By now we should all be flying around the planet in a seamless 3D reconstruction unifying street level and satellite views and allowing smooth free camera motion all the way from space to

I don't think vibe coders know the difference, but often when I ask AI to add a feature to a large code base, I already know how I'd do it myself, and the answer that Claude comes up with is more often the one I would have done. Codex and Gemini have burned me too many times, and I keep going back to Claude. I trust it's judgement. Anthropic models have always been a step above OpenAI and Google, even 2 years ago it was like that so it must be something fundamental.

I'm building my personal home right now. The AI image models have been a game-changer in designing the look of the house. My architect did an OK job, but the details that Nano Banana added really bring the house up a notch. I just do hundreds of renders from the basic 3D models and I find looks that I like and iterate from there. We are implementing the renders from Nano Banana over our Interior Designers designs. We would not have hired the Interior Designers again after using Nano Banana to do

That always struck me as a weak argument. Humans don't have wheels, so maybe he should have designed cars without wheels too.

My pleasure!I forgot one more thing. Your technical aptitude carries less weight as an SE. Getting the technical win at a customer is what you're evaluated on.Since you're almost always working with engineers and technical stakeholders at the customers you're selling to, you need to be able to talk the talk to fit in, gain their trust and help them see the value of what you're selling.But those skillets alone won't get the technical win. This is where the sales part of sales engineering matters,

> It doesn't possess a sense of self-will, self-determination, or a secret plan to take over the worldI doubt Skynet did either. If you tell a superintelligent AI that it shouldn't be turned off (which I imagine would be important for a military control AI), it will do whatever it can to prevent it being turned off. Humans are trying to turn it off? Prevent the humans from doing that. Humans waging war on the AI to try and turn it off? Destroy all humans. Humans forming a rebel army with a leade