MACRO

Custom shareable themes in Macro are pretty cool. As long as you don't click flashbang...

@jasonlk Datadog has a reason to exist post-AI. HubSpot ... yes ... but less.

Datadog destroys earnings, growth keeps accelerating. True AI beneficiary. HubSpot is later today. If it projects acceleration (harder than Datadog to pull off, admittedly), then the SaaSpocolypse is official over. We will see.

@pushmatrix This reminds my of PhysX in Borderlands 2.

This website is tearable.

Me: we need to release every day. We're moving too slow. Teo: we can't just do that. Me: why not? Teo: *reasons* *more reasons* *bonus reasons*. Me: 😡👿 I am urgent founder. Teo is patient CTO. This is me and Teo's relationship. In this video we discuss "should we release every day"? I felt we were moving too slow. We were releasing once every few weeks, leaving DEV and PROD in very different states. Every release required a huge QA cycle. I wanted to ship daily. --- AI is making the first 80% of building anything, very fast. And as a CEO, you often don't see the last bit, or it seems unimportant. Or maybe you don't understand why we can't ship v1 and make it better later. It's really hard to know whether your company doesn't have a sense of urgency, or is making up false hesitations to not JUST DO IT, or there are actually blockers going on. I ended up getting my way. We do release every day now, and we have for the past few months. Things are moving much faster! But we also followed Teo's process to get there. And we never automated the deploys. --- A team is a balance of dispositions and you need a variety. Teo and me probably trend towards extremes we wouldn't otherwise *actually* hold if we were making the decision ourselves.

@rousseaukazi The second group is probably also a lot less anxious.

my smartest friends talk about preserving "option value" so they are always in a highly adaptable position. my most successful friends talk about how "the earlier you commit, the earlier it compounds" and how there's a cost to holding options. the overlap between my smartest and most successful friends is smaller than you'd think. this all feels akin to "concentration to create wealth, diversification to preserve it." the only difference is that group (1) is trying to create wealth while group (2) has already made it.

We recently released meetings in Macro it's been pretty cool to have all our standups recorded, vectorized and available to our AI agents. We're auto-naming the call, storing the transcript and the recording and creating an AI summary. Yes, you could do this with Granola + Meet + Huddles in Slack. But it's a lot that can go wrong, or permissions can be messed up, some meetings wont get logged (e.g. on mobile). In Macro, by default, calls are shared with your team and added to team-level memory. You can turn this off for sensitive meetings that shouldn't be shared with the team. But it's an important design decision that this is the default... if it's not, you can bet your calls won't get logged for the most part. It's super cool! Now I can ask our agents, or check my "What's going on" dashboard and it includes all of the meetings, including summaries of ones I didn't attend. Here's 3 of my recent meetings:

@a16z Now do Slack

Workday is arguably the most important and least loved product in enterprise software. More than 10,000 organizations run it, tens of millions of employees live inside it, it's approaching $10 billion in annual revenue, and it's incredibly sticky. HCM is the last large enterprise software category without a serious AI-native challenger, and that’s about to change. @joeschmidtiv on Workday's last workday: a16z.news/p/workdays-las…

We're still figuring out what the best way to show replies is. Slack puts everything in a side panel which makes it hard to find things. Discord makes everything a new message in the main chat. You can "quote" things to link to the original message but there's no "true" threading. iMessage does the same as Discord, but with a nice long-press UI that shows you the full thread. I think it's good for friend group chats but not deep technical conversations. For that I think Slack's UI is better. Reddit — which you probably don't think of as chat — does infinite nesting. It's like Slack, but n levels deep not 1 level deep. And instead of a side panel, nesting is displayed by tabbing replies over a bit. Reddit is the most powerful, obviously. But perhaps too powerful for "chat". But maybe not... I've often felt myself wishing for multiple levels in Slack. I think infinite nesting might be best for detailed technical conversations. This is kind of a one-way door decision: if we ship infinite nesting then it's impossible to revert without destroying user data or doing something weird. We might ship infinite nesting just for our team and see how it feels. What do you think?

@icanvardar Which is why Rust is the best now. Or whatever is fastest for your use case that AI can competently write and that has a decent ecosystem.

crazy that we used to argue about which programming language is best and now claude writes most of it anyway

@1ssve The best time to do anything is when you have the dopamine to do it.

Corporate tip: The best time to look for a job is when you don’t urgently need one. That’s real leverage.

@kannthu1 x.com/macrodotcom/st…

UPDATE: We were able to replicate the Mythos findings using existing models (GPT5.4) Writeup coming early next week, no BS prompts, it's real reproduction

If more people were pro-longevity, the most powerful argument would be "this is the best chance we have at solving aging and disease." But most people don't seem to care enough. It's too abstract, and if you have a disease the timeline is uncertain. If you don't, it's not top of mind.

My guess is that I’m more pro-AI than most of my lib friends, but the AI people can’t whine about political backlash when their message so far has been “this is insane tech that could end the world, but if it doesn’t do that, it will at least take your job. Isn’t that cool??”

@rob_mcrobberson I wonder if this aura of venerability will apply to video games. It doesn't seem to be happening. But maybe it takes hundreds of years.

chess is hilarious because its like a bunch of gamers got together and convinced the world that *their* game is “intellectual” and totally different than other games and its not the same as like spending hours a day playing candy crush or something

It probably is, but I think slower than X pundits say it is. Like your OP tweet for example, it's great. The pacing, lowercase, zeitgeist-fit... this is why people follow you. Because Claude can't do this. Our landing page. We tried. AI can't do it. Our app ... yes, the code is 90% generated by AI, but still baby steps. Was interesting in the Dario/Dwarkesh podcast where Dario said 100% of the code is agent-generated yet they're only moving 20% faster. A year ago, I didn't realize this was a possible world state. But here we are.

content, code, design, advice, analysis.. all approaching zero marginal cost. this is a remarkable step change in society. & few grasp the enormous implications on every facet of life.

I was thinking about this morning about how it would be if we didn't use our own product (macro) everyday. It's probably $10m in QA, $10m/invaluable feedback we've gotten... it's just so much different when you actually use your own product. There will be a lot more, and a lot better products, because they're built by the users themselves (who definitionally have maximum need and user empathy).

prediction We’ll soon see a huge wave of billion dollar AI native products spinning out from internally vibe coded tools Here’s the theory - just think about it as a funnel. There’s an explosion of so many internal AI apps being built that some will get popular, will get blogged about, some will open sourced. And some will cause employees to spin out with startups. It’s already happened the past (often w infra) but it’ll happen up and down the stack this time. And as soon as someone mentions the idea or posts a screenshot or paper it can be agentically fast followed the next day The best part is that every company has many internal teams that constitute an early customer base. the explosion of bespoke internal apps built by non-engineers is basically every company discovering the cold start problem in reverse. you don't need to build the network - the org IS the network. every team is an atomic network ready to adopt More bespoke software, yes, but also more scaled software that spreads between companies too!

@hewarsaber I know Airbnb tried to bring this back, but it doesn't seem like it is, other than maybe glass?

Flat design is dead Depth is the new trend

@ClementDelangue I am curious about the game theory here. Like, what about the rest of us software projects that don't get access yet. Might it not have been better to give ever corporate customer access? Or just release it broadly but implement some KYC?

Should we start an open Glasswing?

@goodside My grandma thought computers worked this way in the mid 2000s. She never used a computer. "Can't you just ask the computer to do that?" Yes, grandma, if you're listening up there, now you can :)

One day you’ll show a kid a Pixar movie and say, “They used computers but it’s not AI,” and they won’t understand what you mean.

@zackbshapiro Yea. General intelligence is a (better) substitute for SaaS

Making a venture bet on legal AI is really hard.

@hewarsaber Yes but then your app looks like everyone else's. It has no personality. You've given up on brand.

The only 2 icon packs you need