when I'm clicked into a post and then click into that person's profile, I want to go to that point in their timeline. there's usually an additional post around that adds context to whatever the popular tweet is. is that even feasible @nikitabier

it's so tempting to map data engineering onto the vibe coding trend because it looks so similar (it's all just code!). but the reality is that very few people are going to vibe data eng whereas lots of people will vibe software eng.

my god I wish Volkswagen had better software. it's so close yet so far. it doesn't have to be this bad! Tesla has shown the way!

learned yesterday that iCloud supports custom domains (if you're paying for +). way easier than Google if you're already paying.

this post uses 3 different dash sizes. impressive honestly

also how many times you DM'ed a colleague saying "hey quick question" and then waiting for them to reply before asking

Attention is all they need. In 2017, the paper "Attention Is All You Need" introduced the transformer architecture, catalyzing a massive step forward in AI progress. Since the dawn of LLMs, this attention mechanism has become a social narrative and a thesis for rapid ascension in Silicon Valley. The heuristic is simple: attention is all you need. Roy Lee is the most infamous example of a founder who successfully captured attention at an exponential rate. He did it by ragebaiting the shit out of social media and cemented himself as an incredibly polarizing figure. He achieved escape velocity. A16Z preempted Cluely with a $15M Series A after Roy's timeline dominance reached unbreakable highs. People assumed this was now the best strategy for founders looking to be the next Zuck (lol). It was immediately obvious that attention doesn't indiscriminately drive every business objective forward. From the outside, these founders' tendencies seemed bizarre, and they spent more time flexing cars than shipping software; however, attention is all they see. Attention is all their investors see. Attention is all the company sees. There's little regard for customers, the business is always in the back seat, and there is a collapse of self-respect and dignity for those who take this path to its extreme. Does the method really work? It depends entirely on how you're measuring its impact. In a narrative dominated environment, attention is valuable for injecting new trends and attracting capital. These are valuable for institutions looking to control information flow, but within the scope of the company, they’re not very effective for scaling a non-consumer application. I believe there's enough data to conclude that attention is NOT all you need to blitzscale a successful software company in Silicon Valley. After observing the effects of this mind virus for about a year now, I'm disturbed by the results. It's disappointing to witness the derailing of an entire generation of new entrants "striving" to be founders. I don't mean to sound like a boomer, but these kids that are just a couple of years younger than me seem to be programmed with an entirely different incentive structure and model for building a startup. It's now the meta for founders to operate on attention mechanisms, just like they're LLMs. Founders have become incredibly sycophantic, and they will jump at the opportunity to trade dignity and self respect for views. Honestly, I see parallels between some of them and tweakers in the tenderloin. They will stoop to incredible lows to sustain their attention high. Furry costumes, strippers, acting like a cuck, hiring 17 year old girls, etc. Neither they nor their investors seems to have a moral compass. Brain rot startup ideas are rampant, gambling is being integrated into b2b saas, and all roads lead to attention. Real technology isn't being built by this new cohort of founders, and we've entered into an era of hyper competitive degeneracy, fighting for attention. Welcome to the new Hollywood. This time, it’s not cars, but humans that become transformers.

the founder of a company having the last name Shipper who uses AI to ship a ton is another win for nominative determinism

"it's all just data engineering" 🌎🧑‍🚀🔫🧑‍🚀

an upsetting number of gray hairs are making themselves known on my head

such readability from @Apple

can you imagine if apple intelligence was actually good. I desperately need an ai to tell me which of these messages from my kids' school app is actually important... the moms are treating the class chat as their own little group chat...

I'm pretty sure it would go over like a lead balloon, but should chipotle just weigh the food and charge based on that? the froyo place we go to does that and it seems to work!

anecdata but one of the Arlington, TX locations is clean, the staff are nice, and I have no complaints about the portions

I genuinely love that AI agents are pushing people to be more disciplined. It's an outside force that incentivizes more structure behavior. Things that are tough to get people to do for their human colleagues seem easier to get them to do "for the AI".

I read that Harper's article on gooning and damn what a terrible day to be literate

feels like the modern data stack is typified by moderndatastack.xyz - a site that hasn't been updated since 2023, built by a company that was acquired this year by a random AI company. this tweet brought to you by my soon to expire domain - postmoderndatastack.xyz which redirects you to the M$ Excel website

github shaking and vomiting at the sight of this

life is better when you realize you don't have to give a shit about dumb opinions. do what makes you happy!