Building brands in software, apps, services & media. Building. ☕ https://xitter.cf/t.co/Zj0APa4Wd2 🎯 https://xitter.cf/t.co/LWnqQLPgLP 🖊 https://xitter.cf/t.co/j3z7Zk5eGi 📁 https://xitter.cf/t.co/plyPgsop2F
Delray Beach, Florida
Joined May 2009
- Tweets8.5K
- Following405
- Followers3.2K
- Likes25.7K
Waitlist re-opening today.
recap:
1. X had a flash in the pan moment for data brokers. 82% didn't confirm terms, verify person, have deals and are now churned out. Might this wave come again? sure, its X - only takes one AI written, GRQ thread that is total BS, and boom, back to it (at least to some degree).
*I think there was a silver lining here for overall market.
2. much more robust processes in place, much shorter acceptance windows. Geo bound, must verify (stripe identity verification), must accept terms, must list within 30days, must have a transaction with 60... just to name a few.
Chris Riley retweeted
jevon's implications on rl tasks might be somewhat obvious but worth spelling out
as models saturate human work, the surface area for capability & alignment gains starts to shift as the industries transform
take finance: a good task (literally 3 months ago) might be take an earnings call for a stock and update an existing Excel DCF with new mgmt assumptions
model needs taste to "read between the lines" on expectations vs mgmt guides -> you can't take mgmt teams at face value -> the models are now very good at this
finance now moves faster -> pms want live updating dashboards instead of excel workbooks as soon as the mgmt call comes in
consequently, the new surface area shifts:
1. can the model distill the earnings call to the 3 data points that move the stock such that the pm can make a decision faster than her competitors?
2. can the model make a dashboard that is interactive and in-line with house firm style?
3. can the model leverage historical traces and memory to understand how the pm has historically traded this name?
100%; + with the right team and task software (I built this), you can stay on top of the constant shifts on how tasks are continuing to evolve as AI continues to gain adoption across enterprise (large and small).
Chris Riley retweeted
Translating @elonmusk here:
He’s basically saying the easy money for AI data center startups just dried up. Prior to this if you had a pulse, power contracts and GPUs, you could raise gobs of money.
Lenders have stopped giving loans to companies whose only business is building GPU farms (aka neoclouds). Unsurprisingly they now want companies that already make real money from something else. @SpaceXSI can still borrow billions for GPUs because it has @Starlink and rocket revenue. Most of the 200+ pure-play neoclouds don’t have that, so a lot of them are going to run out of cash, fail, or get bought cheap.
Pretty much.
As of the past few weeks, debt markets have almost entirely shut off access to money for any company that doesn’t have substantial revenue independent of AI data centers.
SpaceX closed $40B of debt for GPUs, but we’re investment grade and have Starlink & launch revenue.
That means the 200+ neoclouds are going to see guppy-level death and consolidation. Some will make it through, but not many.
I'll let the legal experts have a take on how this has just become the norm... (sure 1-3 is just some level of fraud)
but no matter what, the majority of the times I think this comes back to bite the business. funding as a moat being a VC darling (on continued fake numbers) is not a strategy I would want to deploy to have a LT, successful business. I think most times... the music stops.
Play long-term games with long-term people
so i have been in the sf tech scene for a couple of months now, and i learned the games that ppl play
1. land a $3M lab order in 2 weeks, *26 = ~$80m as the annual revenue
2. schedule all orders to be signed in the same month, and multiply the best month by 12
3. sign $5M LOIs with friend's companies and batchmates. they sign one back (also happens at scale with large labs/compute)
4. buy github stars at 6 cents each
the last company that swapped invoices with a partner reported $220M - and it's bankrupt now
sad that it's a whole gresham's law game right now and if you dont play you just get downvoted
this is happening more frequently as the market hyper-evolves.
had a business owner yesterday that signed less than 2 weeks ago (probably not even data accepted yet) with micro1, exclusive - but played this card ("I'm not sure the full data is exclusive, I'm really not sure, we will look again"), and are already trying to sell their data again. red flags all over on this one specifically. wide ranging differences from form intake vs. reality.
told them we can't help (even after pressing the exclusive issue and them essentially not caring)... I guess everyone plays their own hand?
Two types of exclusivity in the Company Data biz:
1) Buyer (lab) gets the data exclusively.
2) Broker has the right to market the data exclusively.
Labs dont really care about the former, but the later is super annoying to deal with if the Lab gets the same dataset twice from different sources. If you are repping a dataset to a lab, make sure you have it on exclusive.
business operator (pre AI and AI - truly in the weeds, vast industry experience, ie generalist)... could be a v powerful combo for building tasks.
lets get to work...
pip install --upgrade jax flax optax numpy
Vast AI (Nvidia H100)
jax.vmap
...
cook/learn/cook/fail/cook...
10 minutes.
10 fucking minutes on the computer today..
email 1 from a Vendor: We've been hacked, please rotate keys immediately.
emails 2: Stripe vulnerability; please address when you can.
x: Genetically enhanced wardogs (wtf)
x2: LabCorp was hacked
x3: "Just the beginning, expect things to get worse from here"...
Pup: Rips ...what I believe is his first shart waiting for me to go on a walk.
brb...need to suit up for the day and then clean the floor.
Less than a week to go.
Check your inbox to get verified < 5 minutes with our new broker onboarding.
--> Accept terms
--> Verify your person with Stripe Identification
--> List your first deal
After the 27th, we will reopen founding cohort directly from our waitlist.
Don't forget to reclaim your founding spots by October 27th.
Click the link via email (or just go to your account), go through onboarding, identity verification (via @stripe), and getting your first listing live and awaiting RL desk approval. Looking forward to chatting with you all.
Takes about 5-10 minutes.
obviously the context/harness was likely very important... but... jesus, "find bad vulnerabilities in PHP"... is all it took.
Chris Riley retweeted
Most investors are still massively undervaluing data.
I was also skeptical when I left a data co for Altos in 2025. Today, I’m all in on data.
Every software company will become a data company. Most will lose. Many more will win than we think.
