[reenorap]

The only way AI will be profitable to companies like Anthropic or OpenAI is to make the cost $1000-2000/month or more for coding. Every programmer will be forced to pay for it because it's only a fraction of their salary (in the US anyway) and it's the only way the programmer will be competitive. Whether the company pays for it, or they pay for it themselves, it will need to be paid.

There's no other way that these companies can compete against the likes of Google, and Facebook unless they sell themselves to these companies. With AWS and GCP spending hundreds of billions of dollars per year, there's no way that Anthropic or OpenAI can continue competing unless they make an absurd amount of money and throw that at resources like their own datacenters, etc and they can't do that at $20/month.

[danny_codes]

Even worse, the open weight models are practically indistinguishable from the closed ones. I just don’t see why you’d pay full price to run Claude when you can pay 10x less to run Kimi. There are already loads of inference providers and client layers.

Without heavy collusion or outright legislative fiat (banning open models) I don’t see how Anthropic/OpenAI justify their (alleged) market caps

[leptons]

> the cost $1000-2000/month or more for coding. Every programmer will be forced to pay for it because it's only a fraction of their salary (in the US anyway) and it's the only way the programmer will be competitive.

I routinely match or beat Claude with regards to speed, I often race it to the solution because Claude just takes so long to produce a usable result.

Staying competitive doesn't mean only paying an AI for slop that often takes longer to produce. AI is a convenience, it is not the only way to produce code or even the most cost effective or fastest way. AI code also comes with more risk, and more cognitive load if you actually read and understand everything it wrote. And if you don't then you're a bit foolish to trust it blindly. Many developers are waking up to the reality of using AI, and it's not really living up to the hype.

[reenorap]

You must not be using it right because where I work, a Big Tech company, it's been transformational. Things that would take me a day to code takes minutes. I can't coded since last year. I can see why software engineering as a career is a dead end job now, I spend most of my time testing and code reviewing instead of coding.

[leptons]

So because I can match or beat Claude at the tasks I give it, you think I'm somehow using it wrong?

Maybe you don't recognize someone with real skill and 30+ years of experience? I don't need Claude, but I'm using it. Sometimes it succeeds at simple tasks, but it's out of its depth for anything complex, and after enough iterations on one task, entropy takes hold.

Maybe your coding career was a dead end job, but mine is doing just fine. I'm also not sure you or your colleagues correctly count the time you spend putting into instructing AI vs what you get out that is actually usable. And if you were slow before AI, then I have to ask why you think learning to be a slop-fixer is somehow better than learning how to be a better software engineer.

[reenorap]

I also have 30+ years of experience.

If you are "match or beat Claude at the tasks" you give it, you're using it wrong. You sound like some of my coworkers that are eschewing AI or are minimizing it. The ones such as yourself who find AI annoying or not useful are the ones who are going to go extinct during the next few years.

The new era of programmers aren't going to be the most "skilled" ones but the most mentally agile and flexible ones because things are going to be changing so quickly. No one knows where our field is going to end up but we know the path is going to be fast paced and will keep changing and only those with mental flexibility and agility will be able to keep up.

[leptons]

>If you are "match or beat Claude at the tasks" you give it, you're using it wrong.

Are you purposely mischaracterizing what I said to perpetuate a pointless internet argument? Because it should sounds like that.

I don't sit there all day racing Claude. I do it sometimes just because I'm tired waiting for Claude to do simple shit. It's a benchmark, not a workflow.

>You sound like some of my coworkers that are eschewing AI or are minimizing it. The ones such as yourself who find AI annoying or not useful are the ones who are going to go extinct during the next few years.

You think you know me, but you don't. You're making wild assumptions about me to try to attack me.

This pointless internet interaction is over.

[reenorap]

If your prompt to Claude isn't "This program doesn't seem to work properly. The log files indicate that writing to the database keeps occurring over and over again. What is the cause?" and if it doesn't actual fix the problem without anything more than the above prompt, then you're not using it properly.

[gedy]

The thing is, does it even matter to Big Tech that you did this in an hour vs a day?

[flu_bar]

Amen! I have had this funny feeling of mental atrophy all along. I'm working at a company that wants us to be "ai-native".

The amount of ai slop from diffs and posts is nauseating.