littlecranky67

This. I don't know why OpenAI, Anthropic and others have these high valuations. We have a temporary demand shock on hardware (GPUs, RAM, Storage) but the market dynamics will create more supply in the next years. Once everybody and their mother can get their hands on hardware, we will see tons of token-per-dollar inference services competing against each others. None of these providers will do any training, just hosting open-weights model. Why would anyone give OpenAI and Anthropic their subscription money then.

It is in the article: "the gains showed up when tokens stopped being rationed" - so it is not just a "cost per token before/after" calculation. Once you told your employees "token are free, flatrate for all - go nuts!" they started to use it for all sorts of things and increased productivity. That is not easy to calculate.

Well that is a great idea, and if I were management, I would give everyone a 240€ claude-giftcard-subscription that they can use privately on their private computers. There is however the danger that they would start using their private accounts to fill with company data (which potentially is against GDPR regulations - hence the amazon bedrock deployment).

Funny, but insurances are a good example. Most of what insurances do is administration. Like almost everyone's job there is to sit in front of a computer all day. Lets see how much of that will still be done by humans in 10 years. I think it is one of the field most impacted by layoffs, as they manufacture nothing.

AI will not replace everyone. There will be people steering the AI, and using it. If you are not one of those people, you WILL be laid off. It is like office workers in the 90s thinking they don't need to learn computers and hold on to their typewriters.

But why should pay everyone else for a lazy/inefficient person? Because that what the salary a person becomes is: someone elses money.

It is a shock going through the industry, and we are just starting. I think future software engineers will need more training years and more unpaid time getting skilled. More like surgeons, who need almost 10 years of training until we actually let them operate on live patients.

Given that we are seeing a big retirement wave of the boomers, a lot of seasoned devs will retire in the next 10 years. So there should be enough demand, and a big enough paycheck for that decade long training to make its moneys worth.

The skill of getting familiar with AI use, how to use it, when to use, learning when NOT to use it, is the skill that will land you your next job. Or would you in 2030 expect to hire software devs that never touched AI?

> and that they use their work time for personal projects to ultimately benefit that work time.

That is exactly what I am saying and I would advise them. If you are early 20s right now and a software dev, just waiting for your employer (a large, slow enterprise) to train you an AI seems like a career killer. The whole attitude of "I landed a job with job security, so I am not doing a thing more than I have to" is what drives germany into the ground...

Well I am still in the loop as a senior dev and I get to review everyones code before it gets merged. Official guidelines is that everybody is responsible for the code they produce (with or without AI) and must understand it when they give it to the reviewer. I've been doiing this 3+ years at this company, so before the whole AI thing landed. I am explicitly not allowed to implement features myself (because the idea is the juniors learn and get trained).

Now, within the last 6-12 months of AI advancement, I think you could lay off half or even 3/4 of the team if I was explicitly allowed to implement features myself using AI. But that is not the current policy.

As for worries "ending up on a codebase you can't manage without AI" - my personal take is: No, I am not worried. Because going forward AI will always be available in the development workflow. That is my own opinion, not the corporation standpoint, however.

No you misunderstood my comment. I suggested they spent their private money - 20€ a month for the lowest Max flatrate tier - because to me it is obvious that you absolutely need to develop the skill to work with AI. They only have the "pay-per-token" model on their work accounts and laptops, because that is what large corporations get - no flatrate. That leads them to not work with AI. Most of them just refuse to get that 20€/month subscription to experiment with it on their private laptops. I didn't say private time - company policy is no US cloud services for work related topics so they would have to install it on their private computers. They totally could work on toy projects during their time, as the team is in homeoffice a couple a days per week.

The thing is the have been told to use the tokens wisely, but no limit is communicated. Still everybody is hesitant. To be fair, the company runs claude on AWS bedrock for data privacy requirements at a different pricelevel. But I guess that is the drawback for german job-security: each and all of them know they basically can't get laid off.

I also attribute a lot of it to be just plain outright denial. I have a feeling a lot of people don't WANT to understand that their coding skills they acquired over years have dropped a lot in worth. It is the "as long as I don't step on a scale, I can't be overweight" kind of denial. You are afraid to see what is behind the curtain, so you simply don't do it.

"the gains showed up when tokens stopped being rationed" - I can totally see that. I am leading (as an external consultant) a larger dev team for a huge insurance company. The devs are quite junior or in general more the "I became a sw-eng here because that company is the only big employer in a 1h radius" type of people (meaning: No rockstar developers). Most of them do not play with AI in their spare time, and they have been told to spend claude tokens visely. Everybody is too afraid now to play with it and explore during work hours for fear of being token shamed.

At the same time, most of their junior devwork can be replaced completely by AI - so they really should start to learn how to work with it. I recommended everybody to get a private 20€/month claude subscription and use it for toy projects. I have younger devs (early 20s) in the team that told me they won't spend private money, or simply have no usecase in their private life for AI. We talk about software engineers here, and IMHO most of them are in denial of the fact that their skills have mostly become worthless if they don't start to pivot to something valueable.

This is the final nail in the coffin. I already order my new vservers with debian. I'll keep the existing ones on ubuntu for now, but will not consider ubuntu for anything server side in the future. My reasoning was "Debian is volunteers, they probably won't release fixes as fast as a commercial entity" - obviously the reasoning is backwards.

But I also must say, I am not giving a single dollar to ubuntu nor did I ever do so, so it is also perfectly valid for Ubuntu to not give a damn about me.

How has proton enshittified? Honest question as I am a yearly proton business subscriber.

Yeah I guess I stay with proton. I use drive, vpn, email (own domains), meet and their password manager. Seems I'd have to use individual providers for each of those, and manage a multitude of subscriptions.

> 1) Does expertise atrophy if you are just a human-in-the-loop

Well I guess you use the same technique as for muscle athrophy. I sit all day at work and and drive to places using machines, so my body athrophies due to lack of movement. We built gyms where you go to to fight that athrophy. And we invented jogging and hiking where you go to places without the intent to actually travel somewhere. Same thing applies: You will have to program without AI on projects, that are not your real end product to keep your skills in training.

> 2) How do junior engineers develop the so-called expertise if they have not gone through all the steps and suffered repeatedly?

Same as above. The training + education will take longer, we will have to create training products for the juniors to practice and gain experience. The products will not necessarily be used in production, but should only be seen as sparring rigs.

What do you use as alternative? I came from Google Workspace and Office365, they are horrible. And I don't think Microsoft and Google have a different stance on AI than Proton.

I'm a paying proton customer, I migrated last year from the horror that is Office365 (or Copilot 365 or whatever they call it now) and before that I migrated from Google Workspace. In direct comparisson, proton is the best product for me - and that does not include privacy or security things, just stuff like generall UX and featureset. Yes, they have some annoying bugs, but so did the others.

I think the AI backlash against Proton so far is unwarranted. It is purely speculative and based on make-belief. I am a dev myself, I use AI more and more, and I have no more long-term experience than the next dev. But so far the claim that using AI in your corp and workflow will result in buggier, shitty code is not true on a broad scale. It is certainly not true for proton, at least I don't see service quality detiorating for my service. I think it is a tool, and what you get out of using it still depends on the skill of the craftsman.

As a customer, I do demand stable service and the promised saftey and security. But I also demand low prices and a product that can compete with the market. Now, the market is Google and Microsoft - two companies we don't even need to question if they lay off people due to AI usage. Proton is kind of under pressure to compete on a product level in a pretty much david-against-goliath situation. We see AI-assisted exploits such as the Linux use-after-free from Anthropic and can't blame Proton for restructuring their dev process to incorporate AI to find exactly these kind of security issues. I wouldn't even feel comfortable as a client if proton whipped on a "human code only" label on their services.

Your business competes against other businesses. If hand written code is superior, that business will win.

None.

In the macroscopic view of civilization it is a total outlier, yes.

So you claim that in the last 6000 years of history we had stagnation? I would claim the world economy did nothing but grow on asset-backend currencies - sure, there is some localized, small (as in relative to the timeframe) outliers that don't matter.

Yes you can pick individual outliers, but clearly the world economy has been growing on a gold-based (or other shiny object backed) currency for literally thousands of years all over the world. Picking out individual cases is literally like looking at the 0.1%. Or do you seriously claim we had stagnation in the thousands of years of civilization, and only started to have groth since 1971? Then I know who is the ignorant one.

See also: https://wtfhappenedin1971.com/

It is ridiculous to call for government regulation here. Do you think loss of $3bn will not have an effect? Binance will guaranteed take a look at their security systems and take adequate measures.

That is bs. Before 1971 we had a gold-backed monetary system and no record of deflation in the history of mankind.

The cost of that reversible fiat system is ongoing inflation that is put as a hidden tax on everyone. I’d rather have those 6bn from speculators going to NK if that means we go back to sub-2% inflation for a while

USDT comes in many forms (technical side). You literally can trade USDT as one of the only major stablecoins over bitcoin+lightning network but that is not the entirety of all USDT in circulation. If you look into how this works (because it is fully decentralized running over bitcoin, not some proprietary blockchain) you will find out you can exactly get the market cap, which bitcoin wallet hold how many coins etc. Now lightning distorts that a bit, but market cap is not affected by that.

But it is in the nature of stablecoins that you cannot check their vault. You have no idea if they actually store the assets they claim to store. And whether they will honor their promise to exchange that asset for their tokens.

Example of cargo culting in HN is people claiming that there are no tokens possible in bitcoin (false since taproot assets landed in 2021), not anonymous (false also since lightning layer 2 network existed) or not able to "pay for your coffee" in bitcoin (again, bogus since layer 2 networks). This is all cargo-culted knowledge once learned and true in 2017, never to be revisited.

> People are not so knowledgeable about the other coins because the whole point of magic internet money is that it can't be controlled by anyone

Your problem is treating it as magic internet money. It is not - you literally have the sourcecode to truly assess how it works. Your account is not with some entity. Magic internet money is your bank account - if your bank decides to set it to 0 tomorrow it is "magic" - you have no idea how that happens, and you have no guarantee that it doesn't happen. Or when your credit card is suddenly blocked because your bank decides is "suspicious activity" - again, that is black magic.

Bitcoin is the opposite of magic internet money. It is deterministic to its core. As for other cryptocurrencies - well they are mostly open source so you can check them.