satadru

co-founder @ HireNimbus - agentic marketplace for local service economy. deployed automated ML in enterprises as first 25 @ DataRobot, prev data science @ F100s

[satadru @satadru]

@Grok is so well positioned to dominate the enterprise AI race!

1. zero ipo pressure
2. ⁠unlimited compute advantage
3. ⁠unlimited diverse data across SpaceX, Tesla, X
4. ⁠cursor integration
5. ⁠existing, mature governance and corp structure coming from SpaceX to navigate the growing complexity

[satadru @satadru]

100% agree.

[staysaasy @staysaasy]

If I can make any single recommendation to y'all running startups that are working, reserve an extra 2% of your cap table and dish it out 3-4 years into the company to the early employees who are the smartest, hungriest, and most versatile.

Literally even just 0.1-3% of the company can make a huge difference when it's on top of existing grants and equity plans. These people will stay years longer with those grants in hand. You're basically buying mini-cofounder aggression for much less than the price of a "real" cofounder.

Waiting 3 years is important because you want to see real track record across multiple stages of growth, and by this point you'll be more tired and can benefit from a boost of more drive from surprise additional ownership. You want them to have that Big Drive Equity, that BDE.

Selecting for versatility is the most important part of the equation. You're essentially buying call options on their future ability to be execs of functions that you haven't even invented yet. This is an exceedingly cheap way to buy those options, and critically these options are in the money TODAY, since the recipients are already in-seat doing good work.

[satadru @satadru]

Deeply understanding every user's taste & preferences and then matching with the right supply - is the moat for demand-side agents.

Curating supply upfront is not.

Supply needs to be a massively large universe for sufficient optionality and selection. The job of any personal agent like Muse and Instinct is to "find the needle in the haystack" that perfectly and magically fits the user's need every single time.

[satadru @satadru]

Nothing is a bad initiative at a startup. Let’s try!

But huge risk of sameness.
And producers and consumers are two different things.

The best thing about AI powered recommendations is its ability to read data. Expert intervention may dilute that.

- The expert will recommend the best Sichuan restaurant — but AI will drop it as the user can’t eat spicy food.

- The expert will recommend a must-do during a Maui trip, but AI will drop it as it requires climbing stairs and one of group members has difficulty walking.

- The expert knows a good handyman — AI reads 100+ reviews and knows that handyman is great for carpentry BUT not so good for punch list items.

[satadru @satadru]

Join us at the very first Grok Bot Meetup in Washington, DC. We will do some hands on work with Grok @bot and will discuss how we do things with bots and work along with them.

https://t.co/zVUcwLgiHK

[satadru @satadru]

Grok Bot Meetup: Washington, DC https://t.co/ysDIojsZen

[satadru @satadru]

Grok Bot Meetup: Washington, DC https://t.co/ysDIojsZen

[satadru @satadru]

@sophgoldb It’s possible. They launched during a high travel period and 50% of their transactions are travel related. And they’re not pay fac — they’re “just” running the payment on our behalf. So totally possible.

[satadru @satadru]

Great product with Instinct means continuous usage data ==> subsidized compute for billions of users ==> he needs to raise tons of money ==> investors will give him that money if he 100x/ 1000x their money ==> he needs to do it for money

I don’t think the incentives of customers and shareholders and founders are misaligned here. They’re actually very much aligned.

[satadru @satadru]

@bscholl Ability to get to restroom from any seat without making someone get up

[Ben Pouladian @benitoz]

Nobody's sitting through a 1.5-hour podcast, so a 10-second clip becomes the whole story. That's not fair to the person being interviewed.

Here's the full 3 minutes of the viral Jensen "long division doesn't matter" moment. His point isn't "kids should be dumb." It's that the skill moves up a level. He knew every one of the 200 transistors on his first chip by name. No engineer does that today, and chips are better for it.

You don't need to know how every resistor in an iPhone works to build a great app. You need to see the system.

I'm teaching my kids the same way. They're still learning their math, but the goal is for them to see the forest through the trees: the bigger picture, systems-level thinking.

[Xiaoyin Qu @quxiaoyin]

Prediction: from now on, great product leaders will get $100M comp package, not researchers.

1. 2022-2026 is all about killer models. Right now models still matter, but only for frontier tasks like science, math, trading etc. etc. You either kill cancer or you don't matter.
2. Instead, great products like muse, grokbot make AI truly accessible for people. The proliferation of AI matters way more now.
3. Open weight models make it cheaper and better, and the margin for most models will go down significantly, leading to less profits.
4. Training techniques are copyable, but unique data is not. All things equal(compute, training techniques), unique data is the new oil. Even if your training techniques are second-tier, with unique first-tier data you win.
5. Great products attract sticky users who provide unique data that lead to even better models.

Therefore, product is the one that matters. We are finally entering the gold era of product visionaries.

[satadru @satadru]

@signulll You need a few trips to a handful of offices to understand your blindspots. Start with DocuSign.

It’s not WHAT you do in those offices. It’s HOW you do it. And above all - WHY you do it.

[satadru @satadru]

They are insanely easy to set up and maintain compared to openclaw and hermes.

Meta will use the data for other monetization. If I ask agent to buy something or look for something to buy - but don’t actually buy it, then Meta is likely to push that info to instagram. With Muse, Meta is well positioned on both utility buy and impulse buy. Muse is heavily subsidized. So…

[satadru @satadru]

@SawyerMerritt @bot https://t.co/xXTSi1tizL

[satadru @satadru]

@Vahan55Vahan @atifkhan31 Very cheap. lol.

[satadru @satadru]

Doers are the best thinkers.

[Innovation Council @innovationcncl]

.@JensenHuang DISMANTLES Geoffrey Hinton’s AI doomerism:

“All of his predictions have been WRONG. Enough predictions.”

“Don’t think for a second just because you’re an alarmist that you’re doing a social good.”

[@jason @Jason]

.@JensenHuang is a national treasure 🫡

[satadru @satadru]

I think we need to stop this comparison game that plagued the entire AI world in the last few years. It’s shortlived and often meaningless.

Grok Bot is optimized for work.
Muse is designed for life.

They are aiming to solve two different problems. Muse is extraordinary and it will be touching billions of people. The brand investment alone is a great indicator of what Meta is optimizing for.

[satadru @satadru]

@guilleflorvs And Anthropic is likely to become a bio lab/ pharma co.

[satadru @satadru]

@TPated @austinthesing @NWischoff Huh! You’re one of those. Let’s enjoy her beautiful writing!

She is talking about a growing trend in the dating world in large cities: people are clueless with lots of big requirements - when loving lives revolve around small and serendipitous things.