Most early stage companies are now doing the same thing as well. It's basically a re-invention of the "build fast and test" model that was the norm amongst startups in the late 2000s and early 2010s.
Yes it led to some degree of tech debt, but it also made it easier to experiment, validate, and identify good and bad workflows.
At least in my network, we don't think AI will replace all workers and we strongly believe AI will lead to a significant amount of tech debt, but we do also recognize a lot of work in tech today is busywork.
How is it nonsense? Vision doesn't matter - customer feedback matters.
You start off with a hypothesis (X will solve Y's problem by doing ...), you build a prototype, and then you start testing with multiple Ys. Based on that feedback, you then tweak your initial hypothesis or you scrap it and pivot.
If I showed upper management a functional prototype in a first meeting about a future product, they would assume it was already done and ask when it could ship, while not accepting any dates further out than a month in the future. No way I’d set myself up for failure like that.
The organizations I've worked in and funded would recognize it's a prototype and then staff accordingly unless it a vibecoded portion of a cost center part of the product - those are fine to be one-and-done.
Block signed a friend of mine, they quit their other job, then block was like whoops layoffs including people like this person who hadn’t even started. Super unethical.
Counterintuitively, systems with heavier employment protections can make it more common to cut recent hires.
Employment protections usually come with a probationary period before they kick in, so employers can remove bad hires early. This creates an incentive to remove new hires before their probationary period is up if they're showing any signs they might not be the best candidate for the job.
Even when new hires are good and the company wants to keep them, heavy employment protections favor longer term employees. If the business environment changes and they need to reduce headcount their hands may be tied in ways that require cutting the new hires before the tenured employees. This happens a lot in labor unions, too, where tenured employees have greater standing than new hires when push comes to shove and someone needs to go, regardless of performance.
In Germany we have pretty good employment protections (I think at least!), but this would be legal too. You have a 3 month grace period where the employer can terminate the contract without giving much reason - you gotta survive this period then the protections kick in and they can’t just terminate the contract without a justification and notice period.
It sucks but I think in this case even the best protections won’t help much.
I listened to his podcast episode on the Sequoia podcast a few days ago. Interestingly, his argument was "we don't need middle managers" and he plans to have all 6000 employees eventually report to him.
In other words, companies don't need managers anymore. Except for one manager. Him.
I've come to realize a lot of business trends can be reduced to "higher ups are now convinced that x is not actually necessary".
See "we don't need managers" (flat orgs), "we don't need infra" (DevOps philosophy), "we don't need QA" (devs handling testing), "we don't need product" (product engineering), "we don't need frontend devs" (no code generators) and of course all the AI related workforce reduction.
To me, it says something about how detached leadership is from how the sausage is made.
What is funny is the Adobe executive who proved that yearly employee review process is no longer needed over a decade ago but for some reason the trend never caught on across the corporate world. Seems that some trends just don’t catch on for some reason.
Why does anyone listen to him about product design/buisness running?
Twitter was a success _despite_ him. the original idea was strong enough to blast through all of the odd/wrong decisions he took. The time it took to make hashtags a thing, the terrible scaling, the huge overhiring, and deliberate duplication of teams, and his inability/reluctance to make any product decision. Sure he's got great connections, but he is a terrible leader of a buisness
Most of his product philosophy is negatively correlated with businesses that need to make a profit to survive.
I know what he'll do, he'll have someone make a bunch of agents to manage all these poor people via chat. he'll boast about how AI native the company is, it'll be chaos.
There's no doubt they see AI (or whatever the emerging tech) as disrupting everyone and everything EXCEPT themselves; the more interesting question is: conscious omission or reality distortion field of one?
If every manager initially had 5 reports, a quick geometric series shows that eliminating all managers would save you 20% of headcount. Of course, managers tend to get paid more, so maybe you'd save a larger fraction of wages.
I wonder if that's the main concern or if communication / coordination costs are the larger concern
I met my manager when I joined once, then every 6 or 12 months for performance review (which was aggregated feedback from my peers that he took 2 minutes to talk through: "looks like you're doing fine, if you need anything, my EA can schedule more time").
PMs and Engineers made the prioritization decisions.
If someone was severely underperforming, it'd probably take at least 6 months to notice.
Projects would get shut down with very little notice (though I guess that's been a Google constant).
Within two years they had added 3-4 more layers though, after realizing the managers were, after all, needed.
I've never had close to 140 directs (or even friends) but did get close to 40 (direct reports; never had more than a dozen friends or so). Frankly, it sucks. I was (IMO) doing a terrible job, dropping balls everywhere and not serving the people I had a responsibility & emotional commitment to help. It came down to one of: 1. fail at what you truly believe is your job, 2. give up on what you believe, or 3. don't play the game. I picked #3 and quit, but most go with #2 and many are VPs and CEOs today.
That is insane, Block seems to be very poorly run. The headcount still seems bloated, they'll blame AI and layoff more people for their own incompetence.
After reading a book about the history of Twitter ("Hatching Twitter"), I got the impression that Jack Dorsey is a disturbed individual with a poor grasp on reality. So it's not surprising that Block is poorly run.
I work at a less innovative place, and I see out product managers coming with prototypes, at least solid mock ups rather than just a jira.
They socialize it with potential users, they iterate, they find missing requirements, it's pretty powerful.
The net result is we're building better features faster.
I prefer prototyping to slides. The reason is it helps me understand the problem and edge cases better. Getting AI to build means you could potentially understand it even less than if you put the slides together.
Hiring talent that is passionate about delivering a quality product is more important than ever considering there are so many ways to take shortcuts now that might not be obvious until later.
Can confirm this in my portco's and a couple other peers (one of whom previously founded a major threat intel platform).
If you have product-minded Engineers and engineering-minded PMs, you can merge the two into a single function and remove much of the friction surrounding requirements, prototyping, and launching MVPs.
A couple of these products are already being deployed by F100 security teams as we speak. I also know of one F10 that's building it's own entire security platform from scratch with a team of security engineers working directly with one of the foundation model vendors.
Too many people on HN are divorced or too OOTL from some of these initiatives and then get blindsided during layoffs.
What matters now is DOMAIN EXPERIENCE. Do you understand good development principles and the problem your ICP is trying to solve and how pricing, packaging, and procurement is structured? I don't need a code monkey, process sloths, and queens of the calendar. I need domain experts who can actually execute.
The job of engineers is to engineer - it’s the job of the ‘product person’ to envision what the target market would perceive is much better and figure out a strategy of outwitting competitors.
This product person is a fantasy for the most part - most of them are glorified project managers.
Since the crux of this seems to be about replacing middle managers, what do people think prevents AI from successfully managing 140 direct reports on day to day operations on behalf of a lone CEO? I'm reading "it doesn't work," but that sounds like more of a potential opportunity to me than a truism.
If Block were experiencing rapid productivity improvements from AI why is their flagship Square product still worse than Toast? Toast is eating their lunch day after day.
Prototypes of what? What new products came out of Block in the past six years since pandemic? This makes it sound like Block is a place of innovations when it’s just a rent seeking enterprise.
Is the idea that prototypes give the Permission Granter more fidelity into a proposal and therefore can make better decisions? Whereas before, with Slide Decks, the Permission Granter couldn't experience certain things and therefore couldn't make as good decisions to grant permissions?
So in effect this remains a billionaire figure speaking from their own perspective and we're supposed to care?
Maybe if he had one freaking friend he would realize how effing stupid he has become...
BTW, the easiest way to get fired right now...is to over-use AI in an attempt to fool a domain expert.....or in short do not use it to perform in senior position interviews!
Yes, there is even a compliance post(podcast) about Delve talking about that context aspect of it...
Musk taking over Twitter took a lot of the spotlight off of Dorsey, as though it wasn't already a toxic plaxe. He got a second chance in the public eye to be the visionary that's "one of us" and he's doing his best to blow it
The "prototypes not slides" rule works great for product decisions where the devil is in the interaction details. You can't really argue about a flow in a slide deck — once someone clicks through a prototype, the discussion shifts from opinion to observation.
But I wonder how they handle discussions that are inherently abstract — pricing changes, infrastructure migration plans, org restructuring. Forcing a prototype there would just produce theater. The real insight is probably not "prototypes good, slides bad" but "stop presenting things that should be experienced.
I went to a meeting with a prototype once. It was a single happy path with stubbed data, coded in the most naïve way possible. It was, after all, a prototype just to give a feel for what the interactions would be like.
It put enormous pressure on delivery, since leadership had "already seen it working, how hard could it be to make it to production?"
It's funny (tragicomic) to watch the industry learn the same lessons over and over again (such as "'cheap' overseas outsourcing requires unrealistically precise specs otherwise what would take minutes will take days")
This one sounds like "...and this is precisely why we started using wireframes"
Did the same thing early in my career. Built a quick bootstrap website with like 5 pages and all the data was static. The backend was a year off. It was great for end users but the non-IT managers were dumb. Same issue about seeing something working and expecting the world.
"CEO said a thing" journalism, discussed on HN very recently:
https://karlbode.com/ceo-said-a-thing-journalism/
https://news.ycombinator.com/item?id=47577735
In this case, a CEO is reaffirming their decision to layoff thousands because of AI was the correct decision.
"CEO retroactively justified a thing they did by saying a thing" journalism?
Just "shit". No need to overthink it.
After his stunt with the mass firings "because of AI," employees now bring prototypes, not slides, to meetings with Jack Dorsey.
These clowns live in a dreamworld created by their PAs and cronies
Most early stage companies are now doing the same thing as well. It's basically a re-invention of the "build fast and test" model that was the norm amongst startups in the late 2000s and early 2010s.
Yes it led to some degree of tech debt, but it also made it easier to experiment, validate, and identify good and bad workflows.
At least in my network, we don't think AI will replace all workers and we strongly believe AI will lead to a significant amount of tech debt, but we do also recognize a lot of work in tech today is busywork.
The build fast and test model was nonsense to begin with.
Ultimately it’s a way of saying “I have no vision so I just want to quickly throw a bunch of shit on the wall and see what sticks”.
How is it nonsense? Vision doesn't matter - customer feedback matters.
You start off with a hypothesis (X will solve Y's problem by doing ...), you build a prototype, and then you start testing with multiple Ys. Based on that feedback, you then tweak your initial hypothesis or you scrap it and pivot.
If I showed upper management a functional prototype in a first meeting about a future product, they would assume it was already done and ask when it could ship, while not accepting any dates further out than a month in the future. No way I’d set myself up for failure like that.
Depends on the organization.
The organizations I've worked in and funded would recognize it's a prototype and then staff accordingly unless it a vibecoded portion of a cost center part of the product - those are fine to be one-and-done.
Block signed a friend of mine, they quit their other job, then block was like whoops layoffs including people like this person who hadn’t even started. Super unethical.
I’ve been in IT for 25 years, it has happened to me once, unfortunately it isn’t that uncommon.
In the USA at least sure. This was in a country with lightly better employment protections so it’s quite uncommon.
Counterintuitively, systems with heavier employment protections can make it more common to cut recent hires.
Employment protections usually come with a probationary period before they kick in, so employers can remove bad hires early. This creates an incentive to remove new hires before their probationary period is up if they're showing any signs they might not be the best candidate for the job.
Even when new hires are good and the company wants to keep them, heavy employment protections favor longer term employees. If the business environment changes and they need to reduce headcount their hands may be tied in ways that require cutting the new hires before the tenured employees. This happens a lot in labor unions, too, where tenured employees have greater standing than new hires when push comes to shove and someone needs to go, regardless of performance.
In Germany we have pretty good employment protections (I think at least!), but this would be legal too. You have a 3 month grace period where the employer can terminate the contract without giving much reason - you gotta survive this period then the protections kick in and they can’t just terminate the contract without a justification and notice period. It sucks but I think in this case even the best protections won’t help much.
Only 3 months? I had 7 months in France.
It's usually 6 months probation in Germany, not 3 months
Why did they even hire if they had to just fire a person who hadn't even started. It really reflects to the level of incompetence within the company.
I am sorry for your friend, I hope that he is doing fine, Is there anything that they can legally do for this to block?
That's so messed up. I hope they're doing ok.
I listened to his podcast episode on the Sequoia podcast a few days ago. Interestingly, his argument was "we don't need middle managers" and he plans to have all 6000 employees eventually report to him.
In other words, companies don't need managers anymore. Except for one manager. Him.
I've come to realize a lot of business trends can be reduced to "higher ups are now convinced that x is not actually necessary".
See "we don't need managers" (flat orgs), "we don't need infra" (DevOps philosophy), "we don't need QA" (devs handling testing), "we don't need product" (product engineering), "we don't need frontend devs" (no code generators) and of course all the AI related workforce reduction.
To me, it says something about how detached leadership is from how the sausage is made.
What is funny is the Adobe executive who proved that yearly employee review process is no longer needed over a decade ago but for some reason the trend never caught on across the corporate world. Seems that some trends just don’t catch on for some reason.
Why does anyone listen to him about product design/buisness running?
Twitter was a success _despite_ him. the original idea was strong enough to blast through all of the odd/wrong decisions he took. The time it took to make hashtags a thing, the terrible scaling, the huge overhiring, and deliberate duplication of teams, and his inability/reluctance to make any product decision. Sure he's got great connections, but he is a terrible leader of a buisness
Most of his product philosophy is negatively correlated with businesses that need to make a profit to survive.
I know what he'll do, he'll have someone make a bunch of agents to manage all these poor people via chat. he'll boast about how AI native the company is, it'll be chaos.
There's no doubt they see AI (or whatever the emerging tech) as disrupting everyone and everything EXCEPT themselves; the more interesting question is: conscious omission or reality distortion field of one?
He came to the same conclusion that Steve already had decades ago.
These people are amusing to say the least.
I think Google tried this a while ago (flattening the org).
It didn’t work, so they went back to having managers.
They did, I actually worked there at the time, my manager had 140 directs. It obviously didn't work.
But this time it will work. Because, AI, of course.
If every manager initially had 5 reports, a quick geometric series shows that eliminating all managers would save you 20% of headcount. Of course, managers tend to get paid more, so maybe you'd save a larger fraction of wages.
I wonder if that's the main concern or if communication / coordination costs are the larger concern
Wow, holy smokes… 140 directs. Kind of curious: what did the differences look like on a day to day in that sort of org structure?
I met my manager when I joined once, then every 6 or 12 months for performance review (which was aggregated feedback from my peers that he took 2 minutes to talk through: "looks like you're doing fine, if you need anything, my EA can schedule more time").
PMs and Engineers made the prioritization decisions.
If someone was severely underperforming, it'd probably take at least 6 months to notice.
Projects would get shut down with very little notice (though I guess that's been a Google constant).
Within two years they had added 3-4 more layers though, after realizing the managers were, after all, needed.
I've never had close to 140 directs (or even friends) but did get close to 40 (direct reports; never had more than a dozen friends or so). Frankly, it sucks. I was (IMO) doing a terrible job, dropping balls everywhere and not serving the people I had a responsibility & emotional commitment to help. It came down to one of: 1. fail at what you truly believe is your job, 2. give up on what you believe, or 3. don't play the game. I picked #3 and quit, but most go with #2 and many are VPs and CEOs today.
That is insane, Block seems to be very poorly run. The headcount still seems bloated, they'll blame AI and layoff more people for their own incompetence.
After reading a book about the history of Twitter ("Hatching Twitter"), I got the impression that Jack Dorsey is a disturbed individual with a poor grasp on reality. So it's not surprising that Block is poorly run.
It only works if the person at the top is a real visionary. And you’re talking about a handful of people on earth who have this frankly.
I work at a less innovative place, and I see out product managers coming with prototypes, at least solid mock ups rather than just a jira. They socialize it with potential users, they iterate, they find missing requirements, it's pretty powerful. The net result is we're building better features faster.
How can you be less innovative than Block? Their products are 100% ripoffs of better products
Eh, Square and Cash App were pretty innovative when they came out. The industry is mature enough now that all the products are ripping each other off
I prefer prototyping to slides. The reason is it helps me understand the problem and edge cases better. Getting AI to build means you could potentially understand it even less than if you put the slides together.
Hiring talent that is passionate about delivering a quality product is more important than ever considering there are so many ways to take shortcuts now that might not be obvious until later.
Can confirm this in my portco's and a couple other peers (one of whom previously founded a major threat intel platform).
If you have product-minded Engineers and engineering-minded PMs, you can merge the two into a single function and remove much of the friction surrounding requirements, prototyping, and launching MVPs.
A couple of these products are already being deployed by F100 security teams as we speak. I also know of one F10 that's building it's own entire security platform from scratch with a team of security engineers working directly with one of the foundation model vendors.
Too many people on HN are divorced or too OOTL from some of these initiatives and then get blindsided during layoffs.
What matters now is DOMAIN EXPERIENCE. Do you understand good development principles and the problem your ICP is trying to solve and how pricing, packaging, and procurement is structured? I don't need a code monkey, process sloths, and queens of the calendar. I need domain experts who can actually execute.
What a stupid post.
The job of engineers is to engineer - it’s the job of the ‘product person’ to envision what the target market would perceive is much better and figure out a strategy of outwitting competitors.
This product person is a fantasy for the most part - most of them are glorified project managers.
[flagged]
Since the crux of this seems to be about replacing middle managers, what do people think prevents AI from successfully managing 140 direct reports on day to day operations on behalf of a lone CEO? I'm reading "it doesn't work," but that sounds like more of a potential opportunity to me than a truism.
At face value this seemed cool, but the more I think about it slides or prototypes are the same thing, just a different kind of theater.
I feel like he's just doing it for attention.
Seems like it’s working
If Block were experiencing rapid productivity improvements from AI why is their flagship Square product still worse than Toast? Toast is eating their lunch day after day.
And why is Tidal’s library so much smaller than Spotify’s? And why would I use Cash App if they’re going to try to make it “an interface for AI”?
Sounds like Apple under Steve Jobs.
Prototypes of what? What new products came out of Block in the past six years since pandemic? This makes it sound like Block is a place of innovations when it’s just a rent seeking enterprise.
I'm not sure what the flex is here.
Is the idea that prototypes give the Permission Granter more fidelity into a proposal and therefore can make better decisions? Whereas before, with Slide Decks, the Permission Granter couldn't experience certain things and therefore couldn't make as good decisions to grant permissions?
So in effect this remains a billionaire figure speaking from their own perspective and we're supposed to care?
I have to speak up....
Maybe if he had one freaking friend he would realize how effing stupid he has become...
BTW, the easiest way to get fired right now...is to over-use AI in an attempt to fool a domain expert.....or in short do not use it to perform in senior position interviews!
Yes, there is even a compliance post(podcast) about Delve talking about that context aspect of it...
I bet, considering the massive skill needed for it: "hey claude, turn this presentation into a prototype".
Musk taking over Twitter took a lot of the spotlight off of Dorsey, as though it wasn't already a toxic plaxe. He got a second chance in the public eye to be the visionary that's "one of us" and he's doing his best to blow it
> "I hate the way people use slide presentations instead of thinking," Jobs once said, according to a book published last month by David Pogue.
I wonder what he'll think about these vibecoded prototypes and if it's more thinking or less thinking
"Just two months ago every meeting that we would have, you see a presentation or a Google Doc and we go through it," Dorsey said.
2 months ago they were still using PowerPoints? Jesus no wonder they had to lay so many people off. What the fuck is going on over there?
When I worked at Square 10 years ago it was either Google Slides or occasionally a Keynote presentation. I doubt they've switched to PowerPoint.
[dead]
[dead]
[dead]
The "prototypes not slides" rule works great for product decisions where the devil is in the interaction details. You can't really argue about a flow in a slide deck — once someone clicks through a prototype, the discussion shifts from opinion to observation.
But I wonder how they handle discussions that are inherently abstract — pricing changes, infrastructure migration plans, org restructuring. Forcing a prototype there would just produce theater. The real insight is probably not "prototypes good, slides bad" but "stop presenting things that should be experienced.
I went to a meeting with a prototype once. It was a single happy path with stubbed data, coded in the most naïve way possible. It was, after all, a prototype just to give a feel for what the interactions would be like.
It put enormous pressure on delivery, since leadership had "already seen it working, how hard could it be to make it to production?"
Never again.
It's funny (tragicomic) to watch the industry learn the same lessons over and over again (such as "'cheap' overseas outsourcing requires unrealistically precise specs otherwise what would take minutes will take days")
This one sounds like "...and this is precisely why we started using wireframes"
I wouldn’t say “Never again”, unless you put in the caveat that you shouldn’t do this again for the same leaders.
You still need the slides to communicate risk, cost, delivery estimates etc. a great way to get your own team and move up in the org
Did the same thing early in my career. Built a quick bootstrap website with like 5 pages and all the data was static. The backend was a year off. It was great for end users but the non-IT managers were dumb. Same issue about seeing something working and expecting the world.
> theater
That's exactly what you have to do for the CEO class
Flagged for AI
Or new infrastructure. You bring a demo of a new distributed transaction manager?