Get live statistics and analysis of staysaasy's profile on X / Twitter

OpenAI is finally running the experiment "what if we paid 1/10th the people 10x the money" and I'll admit it does seem to be working

When I first joined my current company (been here for many years) the founders told me two things. The first said - we move really fast, which means you have to get things to a done state fast. Don’t leave open threads because you won’t be able to manage them once you move on to the next thing. The second said - if you see a mess somewhere, clean it up. Anywhere. Literally if the lunch table is dirty clean it up. We have more things to do than people to do it and you just need to fix any problem you see. It’s kind of unbelievable how durable that advice was from a couple of 20-something dudes who just started a business.

Wild how many people claim to be 10x more productive with AI tools and yet I haven’t heard a single person say that one of their coworkers has become 10x more productive.

> 2025: models plateau > 2026: companies stop paying for multiple foundational models > 2026: some company does big article about how they moved to open source model and saved tons of money without losing efficacy > 2027: blood in the streets

I’ve hired like 5 people who have been absolute off the charts in performance and they all had the same background. Coming off of a failed startup(s) where they were in a leadership role. Little other major brands on their resume. Generally quiet/not boastful/humble. The startups made them beasts. The demeanor/lack of other brands made them a little less hunted in market, so we were able to get them. I knew within 2 weeks that each one was a total beast.

95% of why big company people and startup people can't get along is that the optimal move at a startup is to play to win and the optimal move at a big company is to play not to lose. This is a much more important factor than speed, bureaucracy, size, politics, stability, etc

Only took Nikita 2 months.

There are so many companies out there right now that raised like $20M and their CTO is just the cofounder who knew how to spin up an RDS instance and now they have to build real PMF and they’ll never ever find someone to run their Eng team for 1% equity doing 996 and reporting into a 26 year old nepo baby.

So I was skeptical about vibe coding. Everyone was talking about it. Lots of non-coders were telling me they were building stuff. I’ll believe it when I see it I said. So an a guy on our marketing team announces one day that he’s leaving. Why I asked? He vibe coded a tool for BDRs that is doing $80,000 MRR already and growing 20% per month. Guy never coded a day in his life and just got $4M in funding to build it out further. I said I can’t believe it. Neither can I he said. Because it’s not true right? Yup, I said. I just made that up for this tweet and in fact I’ve seen zero non-coders vibe code a business ever in the history of vibe coding or non-coders. Ah man, he said, well if I don’t exist I gotta go. See ya. See ya later buddy.

I really can’t believe how fast Twitter and Instagram and Spotify and Gmail and Slack and Zillow and DoorDash and iOS and YouTube are all developing using AI tools. It’s like every week we get a new feature that is mind blowing and everyone is talking about. Literally having trouble keeping my parents up to date on all the amazing new features they gotta check out. Oh wait wait wait none of that has happened.

A surprising amount of management effort at scale basically boils down to "how much am I willing to compromise the integrity of my organizational structure to make a very high performer happy"

> 2023: we all hate managers!!! > 2024: yeeeeee, no more managers! > 2025: things are disorganized as we’re scaling > 2026: I haven’t had a promotion in three years, nobody coordinates with anyone else, nobody is looking out for me, everything is toxic, people are bullies, we can’t make any decision, leadership vision never gets articulated, this is a mess! Already showing up on AI company Glassdoor profiles.

I got another document full of AI slop this week. 15 pages of heavily AI written slop. AI slop tables. AI slop summaries. AI slop bulleted lists. AI is the greatest spam technology ever created. It gives the incapable the ability to pretend to be capable. It covers their tracks with mountains of shit. The tales of 10x engineers are overstated. The tales of -10x engineers are underreported. Organizations around the globe are slowly being suffocated by new founded prolific output amongst their underperformers. If the doc is wrong it’s your fault for reviewing poorly. It’s the AIs fault for writing poorly. The metrics of performance used to be tied to output volume and quality. Volume is a dead metric. Quality is now much harder to detect. The AI feint of competence is nuanced. Meanwhile AI bots are calling your phone. AI bots are answering your emails. We are slowing down. There are no great improvements to existing apps. There are no new great apps. All of the great new apps are magician tricks. Good enough to buy, nuanced enough to not realize they’re providing no value. We’ve begun our long codependent relationship with the trickiest technology ever invented.

A concise list of advice for new managers: * just learn scrum * don’t put 1:1s at the end of the day * you’re not a therapist * don’t try to be funny until you’re a better manager * go on-call * if you think you have to fire them, fire them * you’re bad at hiring. start getting good at it. * don’t commiserate with your team. gripes go up. * multiple times in your career the limiting factor will be your peer. you need to push them or get their manager to do something about it. * don’t argue diagonally * be healthy * take notes during meetings * learn about CBT * regularly upward report. Ask for help. * never do a low integrity thing because you think that’s how the company does it. * don’t use ChatGPT for writing * say thank you when you give a performance assessment. * if someone thanks you for a raise or promotion, say no thanks needed, you earned it. * you’re gonna make mistakes. don’t let people hold it against you. don’t hold it against yourself. * if more than half your team have a problem, the problem is something you’re doing/you own.

You can literally make a whole career out of being the first person at a startup to start tracking and improving and internal metric instead of just complaining about it. Incidents, support volume, cycle time…there’s somebody who gets a big promo just from being the first person to track and fix it.

Here's why 996 sounds super dumb to those of us who've actually built stuff Working like crazy matters when you're in the steep part of an S curve. 996 (or more extreme versions that I've seen, like 10-12-7) only really matters when there's a huge opportunity that you're in the midst of capturing. It's a tactic to get through your to-do list, it's not a desirable lifestyle. People are viewing 996 as a way to *get to* the steep part of the s curve. But getting into the s curve is one of the main things that you can't force. Also, when you're actually in the steep part of the s curve you don't have time to post and talk about how hard you work. You spend all of your time working and any available free time goes to normal life stuff, like eating, the gym, hanging out with you girlfriend, calling your parents etc. This is why talking about 996 just feels fully performative even though practically speaking many startups do it for regular extended time periods.

Wild how many people claim to be 10x more productive with AI tools and yet I haven’t heard a single person say that one of their coworkers has become 10x more productive.

> 2025: models plateau > 2026: companies stop paying for multiple foundational models > 2026: some company does big article about how they moved to open source model and saved tons of money without losing efficacy > 2027: blood in the streets

When I first joined my current company (been here for many years) the founders told me two things. The first said - we move really fast, which means you have to get things to a done state fast. Don’t leave open threads because you won’t be able to manage them once you move on to the next thing. The second said - if you see a mess somewhere, clean it up. Anywhere. Literally if the lunch table is dirty clean it up. We have more things to do than people to do it and you just need to fix any problem you see. It’s kind of unbelievable how durable that advice was from a couple of 20-something dudes who just started a business.

I’ve hired like 5 people who have been absolute off the charts in performance and they all had the same background. Coming off of a failed startup(s) where they were in a leadership role. Little other major brands on their resume. Generally quiet/not boastful/humble. The startups made them beasts. The demeanor/lack of other brands made them a little less hunted in market, so we were able to get them. I knew within 2 weeks that each one was a total beast.

OpenAI is finally running the experiment "what if we paid 1/10th the people 10x the money" and I'll admit it does seem to be working

So I was skeptical about vibe coding. Everyone was talking about it. Lots of non-coders were telling me they were building stuff. I’ll believe it when I see it I said. So an a guy on our marketing team announces one day that he’s leaving. Why I asked? He vibe coded a tool for BDRs that is doing $80,000 MRR already and growing 20% per month. Guy never coded a day in his life and just got $4M in funding to build it out further. I said I can’t believe it. Neither can I he said. Because it’s not true right? Yup, I said. I just made that up for this tweet and in fact I’ve seen zero non-coders vibe code a business ever in the history of vibe coding or non-coders. Ah man, he said, well if I don’t exist I gotta go. See ya. See ya later buddy.

There are so many companies out there right now that raised like $20M and their CTO is just the cofounder who knew how to spin up an RDS instance and now they have to build real PMF and they’ll never ever find someone to run their Eng team for 1% equity doing 996 and reporting into a 26 year old nepo baby.

Here's what's going to happen in NYC: basically nothing. Mamdani can't raise taxes without state approval, which it won't give. He'll try to do other stuff, and he'll get obstructed every step of the way. The people who say they're going to leave NYC won't. He'll be largely inert and life will go on.

Here's why 996 sounds super dumb to those of us who've actually built stuff Working like crazy matters when you're in the steep part of an S curve. 996 (or more extreme versions that I've seen, like 10-12-7) only really matters when there's a huge opportunity that you're in the midst of capturing. It's a tactic to get through your to-do list, it's not a desirable lifestyle. People are viewing 996 as a way to *get to* the steep part of the s curve. But getting into the s curve is one of the main things that you can't force. Also, when you're actually in the steep part of the s curve you don't have time to post and talk about how hard you work. You spend all of your time working and any available free time goes to normal life stuff, like eating, the gym, hanging out with you girlfriend, calling your parents etc. This is why talking about 996 just feels fully performative even though practically speaking many startups do it for regular extended time periods.

I really can’t believe how fast Twitter and Instagram and Spotify and Gmail and Slack and Zillow and DoorDash and iOS and YouTube are all developing using AI tools. It’s like every week we get a new feature that is mind blowing and everyone is talking about. Literally having trouble keeping my parents up to date on all the amazing new features they gotta check out. Oh wait wait wait none of that has happened.

Is it normal in your 30s to feel a deep and persistent sense of your own mortality? Trying to debug if this sense of urgency (and fear of time getting cut short) is a post-covid thing, a high interest rate thing, a 30s thing, or a personal thing lol.

If AI is helping people code a lot faster and better, why aren’t the products I regularly use getting way better?

A concise list of advice for new managers: * just learn scrum * don’t put 1:1s at the end of the day * you’re not a therapist * don’t try to be funny until you’re a better manager * go on-call * if you think you have to fire them, fire them * you’re bad at hiring. start getting good at it. * don’t commiserate with your team. gripes go up. * multiple times in your career the limiting factor will be your peer. you need to push them or get their manager to do something about it. * don’t argue diagonally * be healthy * take notes during meetings * learn about CBT * regularly upward report. Ask for help. * never do a low integrity thing because you think that’s how the company does it. * don’t use ChatGPT for writing * say thank you when you give a performance assessment. * if someone thanks you for a raise or promotion, say no thanks needed, you earned it. * you’re gonna make mistakes. don’t let people hold it against you. don’t hold it against yourself. * if more than half your team have a problem, the problem is something you’re doing/you own.

Ok I’ve been hard on AI and the fact that major apps haven’t gotten better. But recently I noticed a bunch of smaller apps have gotten way better and integrated a bunch of awesome tech, both AI features and regular features clearly sparked by increases in productivity from AI coding tools. Signal, Dominos, United, Redfin, Starbucks, Chrono24, GChat, Substack have all had major advances in the last year, clearly fueled by and integrated with AI. Hey wait a second I forgot it’s Opposite Day and literally none of those apps have gotten meaningfully better. THERE IS NOT A SINGLE APP ON MY PHONE THAT HAS GOTTEN MATERIALLY BETTER IN THE LAST TWO YEARS. Think about how insane that is. “It makes me a 10x engineer”. I know absolutely 0 people that are outperforming two teams of themself from a year ago. Meanwhile I sit and wade through useless AI meeting notes, and trash AI writing all day long. It is time for us as an industry to say out loud that something is deeply, tremendously broken. We claim that things are faster and better and all we see is magic tricks, little demos that trigger amazement. Do you know what you call something that gives you a quick high and then makes everything worse? We’re hopelessly addicted to the magic of AI and not realizing that it’s destroying then internet, ruining writing and art, and it’s major function is antagonistic, used to replace humans. SkyNet is already live.

Literally every piece of founder advice I see on X is the opposite of what Elon does.

Only took Nikita 2 months.