▶️ exploring ai, building GOLFRiSE, investing in scifi (microfactory, cavalla, digital metal)👽🦾 ⏪ odf1 & early mercury, origin, datanyze, redbeacon; ucsb🚫🎓
california
Joined March 2010
- Tweets8.2K
- Following3.2K
- Followers3.1K
- Likes16.2K
Pinned Tweet
my money is on mercury
giving your agent a card is one thing
but a card is only one part of banking
mercury has invested since day 1 in building the full banking experience
agents will need full banking access to be properly leveraged
👁👁👁 retweeted
it's happening...at scale...all fintechs compete.
Gusto (payroll) -- now does business formation and business bank accounts!
mercury (bank accounts) -- now does payroll, bookkeeping, expense management, treasury
ramp (expense management) -- now does treasury and is becoming a "fintech ai neolab"
I love this. golden age of product expansion and shipping velocity.
customers have more choice than ever.
the fintechs are gonna be alright...it's the incumbents who should be scared.
Fintechs finally have meaningful single digit market share for their core products, but are growing and shipping as fast as ever.
👁👁👁 retweeted
Mercury is on an all time streak
1/
Hiring your first employee is one of the biggest moments for a founder. But payroll, taxes, benefits, and compliance are far more complicated than they need to be.
Now with AI we can make the pain of payroll go away.
@Mercury Payroll* launches today!
👁👁👁 retweeted
Mercury are absolutely MOGGING the big banks on digital banking UX. Nobody is doing digital banking better, and it's not close.
Mercury Payroll is live. You make the hire, Mercury handles the rest.
It’s AI-powered payroll with white-glove compliance support, built into your Mercury account. Try it today: mercury.com/payroll
👁👁👁 retweeted
(banking + cards + books + payroll) x agents = sick
1/
Hiring your first employee is one of the biggest moments for a founder. But payroll, taxes, benefits, and compliance are far more complicated than they need to be.
Now with AI we can make the pain of payroll go away.
@Mercury Payroll* launches today!
👁👁👁 retweeted
@mercury's MCP just saved me hours.
I had it reconcile invoices, card charges, and Stripe deposits. That used to eat up my whole afternoon, but it took seconds.
The best AI use cases are the boring jobs nobody wants to do.
👁👁👁 retweeted
Mercury will be one of the most important companies built.
This is hard to see atm but if you watch closely you will see it.
Excited to announce @Mercury crossing $1bn/yr in revenue and 450k customers.
Just getting started! View article
a billy by the goat
Excited to announce @Mercury crossing $1bn/yr in revenue and 450k customers.
Just getting started! View article
👁👁👁 retweeted
Once arms build arms the hardware gets cheap fast. The money moves to whoever owns the factories and the customers buying what comes out of them
👁👁👁 retweeted
Today we ran a production speedtest, to go from CAD file to finished part in less than a working day -
Digital Metal makes the fastest castings in the west
Topology optimized bottle opener in 17-PH stainless steel
Cast by @digitalmetalinc
👁👁👁 retweeted
We can borrow a mechanism from nature: an organism made of cells that can produce new cells. These cells become building blocks for growth or for creating a new organism.
Conveniently, we already have robotic cells capable of electronics assembly.
An AI agent coordinates a group of them to build and deploy new cells. It specializes each cell with fixtures and tools for specific tasks such as assembly, calibration, testing, 3D printing, etc. Together, these cells form a self-replicating robotic organism/factory.
Sounds like sci-fi, but recent releases of Astra/Opus can coordinate robots and control them by writing reliable code. That makes this already possible.
the self-replicating factory
a major inflection point comes when the robot arms start building robot arms
👁👁👁 retweeted
This is the gpt3 moment of robotics
Over the past month, we completely switched to frontier LLMs to control a robot in a customer deployment.
By doing this, we have a classic System 1/System 2 architecture, but on steroids.
For the high-level reasoning layer, we use an agent based on Opus and Astra, which are many generations more capable than any VLA. For the fast execution layer, we use computer vision code written by the model, which is fast, editable, and reliable.
As a bonus, we got in-context learning, which is just a basic feature of LLMs, and continual learning, since whenever a new situation arises, the large model steps in and adapts the pipeline.
And it just feels better to be on the right side of the Bitter Lesson, benefiting from the billions being invested in frontier models around the world.
👁👁👁 retweeted
Can confirm that flying cars do not interfere with The Ocean Cleanup's work whatsoever.
👁👁👁 retweeted
we're launching actions on @mercury MCP
you can now categorize transactions, queue up transfers or payments, and a lot more coming
we think you should be able to have your agents running with your banking, invoicing, spend, accounting and payroll
1/ Starting today you can take actions on your @Mercury business and personal accounts right from your agent (ChatGPT/Claude/etc) via the MCP.
Our MCP already let you ask questions about your accounts.* Now it can move money too (with your approval).