Hacker Newsnew | past | comments | ask | show | jobs | submit | htrp's commentslogin

Founding Eng roles will ask you to take a cash haircut in return for equity.

absolutely be skeptical


>we're not making this decision because we're in trouble. our business is strong. gross profit continues to grow, we continue to serve more and more customers, and profitability is improving. but something has changed. we're already seeing that the intelligence tools we’re creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company. and that's accelerating rapidly.

This is one way of making an all-in bet on AI.

>we're not going to just disappear people from slack and email and pretend they were never here. communication channels will stay open through thursday evening (pacific) so everyone can say goodbye properly, and share whatever you wish. i'll also be hosting a live video session to thank everyone at 3:35pm pacific. i know doing it this way might feel awkward. i'd rather it feel awkward and human than efficient and cold.

Well that's interesting, wonder if we'll actually get a proper accounting of which departments take which cuts.


Even if the AI piece isn't really true - smaller flatter teams will move faster anyway. I always wonder having worked in a lot of startups with 10-50ppl, what on earth a business does with 10000.

> I always wonder having worked in a lot of startups with 10-50ppl, what on earth a business does with 10000.

If a small business needs to send a replacement widget to a customer in a foreign country, they label it "$0 value" (as it's a free replacement part) and mail it with a swipe of a corporate credit card.

If a large business needs to do the same thing, the sender asks the mail room, giving them a budget code and delivery address; the mail room contacts the widget designer for a HTS code, size and weight; then contacts their shipping broker for a quote; then contacts the finance department to raise a purchase order; the finance department contacts the budget code owner for spend approval; then raises a purchase order; then forwards it to the sender who forwards it to the post room who forwards it to the shipping broker who arrange a collection. Later the shipping broker will send the post room an invoice against the purchase order, which they'll send on to finance, who'll query the sender who'll approve paying the invoice.

> Even if the AI piece isn't really true - smaller flatter teams will move faster anyway.

Quite possibly - but you have to remember to remove the bureaucracy, not just remove the people who operate the bureaucracy. If you try to do the large business process with the small business team, it'll be even slower.


Seconded. My experience has been that -- even while still complying with lots of overhead (e.g. government regulations and compliance) -- smaller teams of 1-3 devs move waaaaay faster than teams of 4-10. Could definitely speak to the overall codebase quality or some other factor, but yeah.

Brooks explains this in The Mythical Man Month when he discusses how adding people to a delayed project increases the delay. Communication complexity grows exponentially as team size increases. It quickly reaches the point where it has to be controlled by procedures, forms, approvals, etc.

I expect it's more that early in projects you move faster, and that normally involves fewer people.

Once projects get bigger they need more devs and also move slower.

Put a team of 1-3 devs on MS Word and see how fast they move...


I found this an interesting question and did some research out of curiosity

[Full credits to wikipedia]: https://en.wikipedia.org/wiki/Star_Division (The company behind what's gonna be essentially StarOffice/Later OpenOffice/Libreoffice given Libreoffice is a fork of OpenOffice)

Star Division was a German software company best known for developing StarOffice, a proprietary office suite. The company was founded in 1985 by 16-year-old Marco Börries in Lüneburg, and initially operated as a small startup. Its first product was StarWriter, a word processor that later evolved into the StarOffice suite.

Their number of employees by the late 1997/1990's from the wiki article suggests 170. They/StarOffice achieved over 25 million sales worldwide and held an estimated 25% share of the office suite market in Germany by the late 1990s

There aren't many true MSword alternatives for what its worth but I found a gnome project which is interesting from alternativeto https://gitlab.gnome.org/World/AbiWord/-/project_members

There seem to be 5 main members (I am not counting the Gitlab Admin and administrator)

Interestingly, If I remember correctly, I saw Alexandar Franke in here, I have actually talked to alexandar franke a long time ago on matrix back when I used to use fractal. It was definitely a fun surprise to see him in this project as well.

Aside from that, I think the problem with MS word to me feels like it tried to copy the features of previous word processors including quirks and now anything which wants to be MS word competitor is sometimes forced to copy these quirks as well which to me feels like the stressful cause for the reason why we don't see too many new approaches within this space (in my limited opinion)


Star Office was really good in the 90s too

No it wasn't. Are you perhaps thinking of WordStar?

Yeah what I really meant is imagine if you reduced the actual MS Word team to just 3 people. They would not move fast because by this point Word is an enormous mature project and they wouldn't even be able to touch 1% of it.

Also AbiWord is dead, sadly.


Depends. Do those teams’ code interact with any other code from any other teams at the company?

Every business metrics needs people to safeguard. That's how you get the number of ppl.

Sure but it'll still be a 6000+ team - I doubt nimbleness will occur now.

They're still a megacorp, roughly, with like 6k people remaining. That's a huge company. Huge companies need hierarchy to function, the "flat" thing is a really dumb idea. There's no way to make it analogous to that <50ppl team that executes well and moves fast. To do that you actually need to have a small company.

First you take a 50 person org. Then (for scale) you hire highly motivated performers who, because they came up in big orgs, are used to using 50 people for three years to do a project six people can do in three to six months. Then you create incentives that make them compete for standing. And the standing also depends on their personal scope (ie headcount).

Their shareholder meeting is later today. Maybe we'll find out.

> i'd rather it feel awkward and human than efficient and cold.

So deeply ironic considering he claims he’s doing this because AI can do the jobs these people did.

These billionaires will learn one day that removing humans doesn’t stop at the bottom layer. It’ll continue to happen at layers above until their own position starts to be put into question. They’ll realize those people who are removed due to AI taking their jobs still need to put food on their tables. It’ll take time, but ultimately there are only so many ways that can go. The answer will be extreme taxation on the billionaires.


I do genuinely wonder about the endgame here. Why would the objective winners of the _current_ system, our billionaire class, want to disrupt that system? Do they really believe that they will necessarily be winners in the new world too, are they that arrogant?

They already understand the current system and status quo is going away. They understand, on some level, the consequences of the technocapitalist system they've built and perpetuated.

They're making their own accommodations, rather than trying to change the course: https://www.theguardian.com/news/2022/sep/04/super-rich-prep...


I think assuming human agency (building technocapitalism, correcting course) or the possibility to escape capitalism and its consequences (in bunkers), underestimates what capitalism is.

I question how much of this is really AI vs them just regrouping around their core products and shutting down a lot of ventures or tertiary projects. Either way, the messaging we're seeing is a real shift from the ZIRP ear. Tech companies used to use headcount as a metric of growth. They'd be hiring just to say they're hiring because it looks like growth. Now it's in vogue to boast about your AI adoption and how many fewer heads you need to operate. I think both are lot of blowing smoke, but now it's going to hurt a lot of people.

> but something has changed

i.e. we finally decided to audit head count from post covid-era.

> paired with smaller and flatter teams

i.e. management was axed


you don't think LLM impacts on productivity were a factor at all?

If LLMs really multiply productivity, why would you fire people and handicap the boost?

I have 100 people that can now do the work of 200 people thanks to a new tool.

How is the logical response to fire half of them and bring my productivity back to where it was before?


Because there isn't an unlimited amount of productive work to be done. Sure, a bowling ball factory in a world that demands unlimited bowling balls should take the productivity multiplier AND retain the employees, because they ought to make all the bowling balls they possibly can.

But CashApp jira tickets are not a bowling ball factory in a world with unlimited bowling ball demand. At a certain point, you're just paying people to sit around, or even worse, pretend they're busy.


That’s my point. The letter claims this is a decision made for the purpose of growth, which makes no sense.

This is admitting the company is in maintenance mode at best


> If LLMs really multiply productivity, why would you fire people and handicap the boost?

Presumably, because some of these areas are cost centers versus profit generating.


He explains the rationale, smaller teams work faster.

we're already seeing that the intelligence tools we’re creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company. and that's accelerating rapidly.


This is just rephrasing the same concept.

Claiming than a small group with AI can accomplish more than a large group with AI doesn’t make sense.

More likely the company doesn’t have enough work for the large group.


Have you worked at a big company? It makes sense to me that a small group would be much more productive than a large group, even without AI. Throw in some AI help, and it could be much better.

> It makes sense to me that a small group would be much more productive than a large group

That's not the scenario. The scenario is a large group vs a large group cut into a small group.

The chaos and disruption of slicing 1/2 the company would more than offset any gains. We got people. Not machines. Not everyone adapts so fast. Team work and efficiencies take time.


I do see fewer Square terminals these days, more Toast (and other options too I think).

Demand inelasticity.

> our business is strong. gross profit continues to grow, we continue to serve more and more customers

I would say the vast majority of people in this thread don't believe that this is related to AI at all, other than as a pretext. It's kind of incredible.

As in AI is a belief? This isn't a religion. IF things don't add up they don't.

What are you thinking about for the new best practices for software engineering?

That's the question I'm hoping to answer as I write the rest of this guide, which I expect to take several months: https://simonwillison.net/2026/Feb/23/agentic-engineering-pa...

I'm not sure anyone has a confident answer to that yet though - I certainly don't.


Simon, I'd be interested in a voice conversation with you on this topic. I've developed my own chain-of-thought system that is very different from what others are doing. Mine is a Socrate Agent that leads the developer, does not write their code, but guides them to become a better developer. It completely opposite the direction the larger industry seems to be going. Augmented synergy between the AI and developer, not the AI coding and the human overseeing.

thats almost always been the case with 3rd party human task services

Yup. I was surprised that the article author took the results at face value. Having results that match the most commonly-known AI platform's results perfectly seemed worthy of a mention!

Granite is not very good

Oracle is trying (and mostly failing) at frontier model training


"After all, if Dario Amodei had bought puts on IBM, and the dozens of companies that have plunged more than double digits in recent weeks, he would have made billions, certainly enough to fund his company for months if not years. "

Anthropic put out another blog post about modernizing/migrating away from COBOL several months ago IIRC, it is surprising that this was not priced in already

Feels like a rip from Matt Levine who has been saying the same thing for a while now.

It's not so original an idea that we need to be worrying about who came up with it, I think.

Next up:Epic and MUMPS

dark factory

it's unfortunately not just your algorithm, but the views and likes of people who match your demographic specs....

literally or figuratively?

Google Wave was launched, marketed, and killed in 2 years

Is this supposed to be evidence that Gemini is mature?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: