People seem very attached - emotionally, almost morally - to a model of hiring that looks like this:
* The company decides to open a specific position, which it will fill no matter what.
* The position will pay the same amount no matter who fills it.
* There exists a binary condition called "qualified" which people metaphysically are or aren't.
* The job must not go to an unqualified person, but could go to any qualified applicant.
* Society may instruct you to use certain non-merit-based criteria to prefer some qualified applicants over others, but hiring managers and teams must not use their own subjective feelings or relationships to do so.
I can understand how a civil service or other massive bureaucracy full of principal-agent problems arrives at these rules, but they are obviously not the only valid way to assemble a high-performing team or the best one in all circumstances.
People conservative enough to drop the required nuclear bomb on the transit workers union to be able to deploy something like this won't pay for it, and vice versea.
Existing drivers could be moved to customer service roles on the trains (driver+conductor on 1 train -> one coductor on 2 trains), or moved to non-autoamtable roles (buses, trams).
True, but if you read the comment as U.S. specific, it makes perfect sense. The Democrats in the U.S. are very beholden to the unions and refuse to do anything that harms the union even if it would help the vast majority of people much more. Meanwhile, the Republicans hate transit and while they hate unions too, they hate transit even more and so they're going to not do anything that would help transit. If you're a transit advocate, you have neither party that was actually going to serve you. One party actively hates you and the other party gives lip service but they like other things more and will do things that are harmful to transit because they help the other things they like.
I don't know how to solve this problem, but making people aware of it is a good first step. It helps people who are in favor of transit yell at all politicians who are not helping them. Blind party allegiance is never a good thing.
I think the US problem is that unionized transit workers (and a dwindling number of other unionized workers) can do quite well, but if they were to lose that job, they wouldn't be able to get anywhere near the same salary, benefits, etc. in another job with similar duties. I don't think this is as true in other countries with stronger labor protections overall and a more equitable healthcare system.
Also, the culture for whatever reason seems to emphasize enabling gamesmanship around overtime, special duty pay, etc. as opposed to paying a good steady rate for a normal day's work.
>Meanwhile, the Republicans hate transit and while they hate unions too,v
People hate unions too when they're directly affected by them. Case in point, the dock workers strike from a while back. When people realized their Chinese made slop Amazon christmas presents would be delayed because of the dock workers strike, they suddenly stopped empathizing with them and hating them for it.
Inter-class solidarity isn't much of a thing. If one worker class starts breaking things in society/the economy to get a better deal for themselves at the expense of the other classes, even if they're on the same level, the others will hate them for it.
Another example would be if airline workers would go on strike for better pay and working conditions ... during summer. Imagine how many people who paid for holidays would empathize with them. We are an individualist society.
A chat window that takes multiple seconds to respond and behaves non-determinstically is a pretty bad paradigm for most computer tasks. It just happens to work for software development in particular.
Twitter consensus now is more like LLMs will be entirely commoditized, apps & features will be cheap, and only systems of record will have lasting value.
A chat window that takes multiple seconds to respond, versus your buddy Guy who just went out on lunch.
Non-deterministic behavior from your chatbot, like how the visual design artist's client keeps sending back and forth emails asking for it to "pop" more
You ask the chat window to build a new tool or visualization or simulation or data editing tool for you...and it spins for a few seconds or 10s of seconds then presents it to you, ready to be re-used and designed to interact with both you directly and the AI itself.
Perhaps even by other people, and over time, continuously improving as we discover new use cases or that things are more complicated than they seemed on the first pass. Like... regular software, and not "do[ing] it all in Claude or Gemini or ChatGPT"
Still shocked that people use LLMs as automation for some things that should be done with software.
“Claude deploy this to dev”
vs
“Claude build me a tool that allows me to deploy to dev with a single click / command from CLI”
Orders of magnitude less expensive for the latter than the former. The former usually costs a few cents and the latter maybe a one time cost of 40 cents and then free forever afterwards.
“Claude make this thingy run on our test thingy so I can take one final look before I let our users access it. And make a button for it so I don’t have to ask you again.”
"Claude, read all my emails and make an app that satisfies my customers."
"Claude, find an underserved niche of SaaS software where people aren't yet interested in using AI; generate a customer base through social engineering; and serve that customer base an app that they want. Here's my bank routing number."
"Claude do business."
"Claude build a nuclear reactor."
Your argument is thought-terminating. There's no way forward in this discussion if you believe that AI is capable of literally anything, and there's also no point in having a discussion in the first place if that's true because we're seconds away from singularity.
Yeah for me the accelerator has been using LLMs to build engines, not apps. Build something that outlasts the LLM, or could even be trivially used by a local LLM.
I'm under no illusion that we'll have cheap LLM access for much longer so I'm hedging my bets.
Claude deploy to dev will use abstractions to do it with software , even more it will deploy you a trigger to do it automatically . Otherwise . Nothing wrong of doing it with voice / text request vs navigating to some ugly Jenkins , or half dead github actions etc .
> Claude deploy to dev will use abstractions to do it with software , even more it will deploy you a trigger to do it automatically
Who’s building the system that does that? Do you believe every request that we make - in every context that we make it - Claude will be able to intuit whether it’s something we only want to do once and is then just throw-away, or do it multiple times so a system should be built to facilitate that?
Even if Claude is somehow clairvoyant enough to know that, it still costs money to interpret my request in the first place. Why would I spend 5 cents to do something I could do for free? Maybe once, but across the day I’m doing hundreds or thousands of things and the money adds up. I’d rather have that money in my salary rather than having my employer pay for an AI subscription.
What’s the advantage of doing these things through a chat interface, aside from lining VC pockets? Vs clicking a button or invoking a CLI command? Just for the aesthetic of relinquishing control?
> Nothing wrong of doing it with voice / text request vs navigating to some ugly Jenkins , or half dead github actions etc .
I’ve automated away most of my daily browser activity using LLMs to generate tools. It’s become a lot easier recently. I do almost everything from a terminal, which makes it incredibly easy to e.g. pipe content into files for my LLMs to work with, like errors and stack traces. I understand some people have given access to their entire desktop to LLMs but that seems very solution-specific and I don’t want to commit days or even weeks getting my system to work with one piece of AI software just for another one to win the battle in the next few months. Too much churn. Text and terminal are evergreen.
What kind of browsing activities did you automate away?
If I think about what I do, like reading HN, reading some subreddits, watching series, asking LLMs things, ... I don't see how any of it can be usefully automated away, in such a way that it achieves an equivalent result.
So at the end of the day you still need an app… and maybe instead of spending tokens every time you want an app you just use a prebuilt one… :surprised-pikachu:
If you just need a narrow subset of the functionality for now, might be cheaper in tokens to just vibe it than the attentional cost of searching for just the right software...let alone paying for it.
Why people bringing token prices ? With my Claude subscription I can do everything from building apps every day to doing room planning and science researches , for a fixed price. Software is not cheaper . Before you could argue oh but it’s a subsidy etc , in reality you can do same things with Chinese models just half a year away for also cost of coffee . It’s cheaper now than sales tax you pay for your saas
Enterprises needs to pay for API, they cannot use subs (for various reasons). In work environment it’s a higher chance you will pay raw token prices than have a sub.
Sure , I’ll share you my example I was doing home planning from a shitty plan I had from previous owner . Claude created that with a fraction on my quotas . Eg ~1 dollar . It converted a shitty drawing into 3d , modified the room I wanted , created a few options of visuals, and shopping list from my local store with budgets . Oh and yes it create a local html page for my problem it even created a name nook, not sure why .
With regular software I need an entire package of apps, a specially trained people , and a budget of at least $10k for the same service . Will my services be such a commodity very soon , sure .
Do I want to pay $10k now for manual software and people . Not really , what is the point ? Same way as accountant will not pay one day for the software , I will not pay adobe and designer today .
Have you tried chatjimmy? It operates at the speed of thought. Imagine this with one of the recent frontier AI. AI that operates faster than you can express your thoughts seems like a foregone conclusion eventually but may not make sense right now as they try to reach the limit of what they can make in a model.
"speed of thought" is wildly underselling how fast chatjimmy is. The speeds claimed by system one decision models are slow in comparison to chatjimmy, which produces actual text.
A chat window that will deploy you perfect ui for any problem you have based on your personal taste , that is funny enough is not that different yet. And yes , that ui will open in milliseconds , and it will be deterministic when needed and agentic and smart when not . This is already happening on a small scale with Hermes . With models getting smarter it will
Only get better
you use the chat window to make the software which is deterministic and can be tested. Then the chat window uses that software to do the analytics or whatever else you need. Op is correct.
99.999999% of people have no interest in making software, they just want to do whatever they need to with minimal hassle. You may as well say restaurants are going out of business because microwaves exist.
Even the people making software barely want to make software honestly, but you gotta make that daily bread somehow and sometimes it's kinda cool I guess? But like people overestimate the desire of the average person to be their own software maintainer, most people consume, not produce, regardless of ease and trying to turn everyone into a producer is... kind of the opposite of where society has been going for a very long time.
There are areas of expertise or experience, where I know, that no one else has actually made the right tool for me, so I have to make it myself, and I like being able to do it myself. One is a good tool for learning the vocabulary of Mandarin. I just have very specific ideas how it should work and look and I don't think there is any other tool, that does it like my personal tool. I like such projects, but I think even among developers, the percentage of devs thinking like this is decreasing, as there are fewer and fewer people in this profession, who are in it because they actually love developing things, compared to more and more who do it for the money.
Everyone is saying this about "the factory." I have seen some successful fix and migrate campaigns for well-defined, local transformations such as retroactive fixes for new lint rules, flaky tests, or library migrations. But so far the attempts to do nontrivial new functionality in this manner have made big messes that take more effort to review and correct than simply hand-feeding 1-3 vanilla Claude Code sessions with my full attention.
That's the economy. It's not like all our problems are solved or everything that needs building is finished, even in very AI-forward tech companies. We are busier and working harder and longer hours than ever, and we've been hiring thousands of people in India.
That's true too: AI currently doesn't live up to the promises that were made to the execs, so there's also more pressure for human slop and outsourcing to fill the gaps.
Essentially every high schooler had an active Facebook account and essentially every school paper was written on a word processor by, like, 2010. By 2018 even Instagram is starting to be played out. I would be very surprised to see a material increase in the share of US or western students using social media over the 2018-2022 period.
Feels like a hobbyhorse. The article discusses issues between the educator and the child, like:
> An even bigger problem, though, is muddle-headed thinking among educators. Some have fallen for the leftist canard that tests are harmful and grades are racist. Some have allowed reasonable worries about children’s mental health to warp into an excuse for low expectations. Some unhelpful habits acquired during the pandemic have stuck: children are still missing more lessons than before, and many teachers who relaxed their grading standards have failed to tighten up again.
I think we should look there and focus on what’s actually happening between the educator and the child.
Like so many conservative scare tactics, this is complete bullshit.
I hang out with a lot of teacher. Grade inflation is because if you dare give anything less to an A to a student, their parents complain to administration and administration always gives in and never stands up for the teacher. Same if students aren't turning work in - parents complain, administration says to pass them.
Educators are underwater, overworked, burnt out, and it's getting worse - and none of that is their fault. Every year they're asked to do more with less, which is why many of the best have retired or moved to other work.
I think it can shift rapidly when an addiction or compulsory behaviour becomes a problem.
For example, made-up numbers, but: 1-2 hours of doom-scrolling daily is okay, as it still leaves time for 1-2 hours of schoolwork at home. 3-4 hours of such doom-scrolling, doom-scrolling between classes or during classes will have a measurable negative effect on your academic knowledge and abilities. (I don't have much information on how US schoolkids live today.)
And surely one could never predict the impact of sending kids home for two years and expecting them to manage their own education on their own honor. Surely school aged children are better at paying attention to Zoom meetings than I am as an adult. It's in the computer! Kids are good at computer!
I'd say the main difference is that you couldn't take a device with good media capabilities, good and cheap-ish networking and lasting battery everywhere. Laptops, netbooks, tablets were a tad too bulky, lasted only a few hours unplugged and mainly wifi, and mobile internet was not all that great.
True leading up to 2010. The couple of kids who had the original iPhone or the 3G were unusual. Most of my classmates - including myself - got one between the 4 (2010) and the 5S (2013).
People need to stop believing insane, delusional things like the existence of a human being with a certain name, address, phone number, birthday, SSN being secret or private information.
Downstream of that, people need to stop accepting knowledge of the basic public metadata fields or possession of images containing them as evidence of identity verification. Do actual public key cryptography on the internet or check biometrics and the document’s physical security measures in person.
I do not buy that LLMs and the capability to run them are fundamentally different from other software or general purpose computing infrastructure in this regard. Moves to ban open source software or force OEMs to put little cops in everyone's computers are bad.
The proposal I’m hearing is to make people training models accountable for anything users do with them.
Obviously you’re going to keep the model behind an API and be very selective about the people allowed to call and the queries it’s willing to answer, in that case. As Anthropic has done with Fable. But that is voluntary restraint - mostly in today’s regime we get frontier capabilities in open weight models on a ~year delay.
It’s to keep “the people running it accountable.” The context is agents ran by OpenAI/Anthropic doing damage outside.
There is no user-involved damage. No one is recklessly running agents by the thousands without air-gapped containers, except “the people” than run these labs.
One through eight read as the list of sacrifices you make in having children. But the experiences of loving and being loved back seem at least incomplete without that.
reply