Automation Anywhere

Aakash Gupta

@aakashgupta

The number to pay attention to is $650. That’s the cost of the Tesla AI4 chip this system runs on. A single UiPath bot license costs $420 to $750 per month. Automation Anywhere charges $500/month per unattended bot. The global RPA market hit $28 billion last year, and the entire industry is built on software robots that follow scripted rules and break the moment a UI changes. What Elon is describing is a system that watches a screen in real time, understands what it’s looking at, and acts. No brittle scripts. No custom API integrations per application. The architecture splits the work: Tesla’s component processes the last 5 seconds of screen video and keyboard/mouse inputs at speed (System 1), while Grok reasons about what to do next (System 2). One layer reacts. The other layer thinks. That split matters more than people realize. Every other AI agent routes both perception and reasoning through cloud-based LLMs. That means latency on every action, and cost per inference call that scales linearly with usage. Running perception locally on a $650 chip while reserving the expensive Nvidia inference for reasoning means the cost curve bends differently. The cheap part of the job stays cheap. The expensive part gets called only when needed. SpaceX acquired xAI last month for roughly $1.25 trillion combined. Tesla invested $2 billion in xAI’s Series E. The two companies now share hardware roadmaps, training data pipelines, and inference infrastructure. Tesla’s real-world computer vision expertise from 7+ billion miles of FSD data is now connected to xAI’s reasoning layer. That combination of visual processing at the edge plus centralized intelligence is the same architecture Tesla uses for self-driving, now pointed at a computer screen instead of a road. The RPA industry charges per bot, per month, for automation that breaks when someone moves a button. If this system works at the reliability threshold enterprises require, a $650 chip replaces a software licensing model worth tens of billions annually. That’s why he named it after Microsoft.

Audit The Herd

@AuditTheHerd

This is interesting…. Automation Anywhere (2nd to $PATH in RPA) is looking at a reverse listing (last valued at ~$6.8B ) acquiring the publicly traded C3 AI (~$1.8B market cap). https://t.co/PMrE9wzvEZ

Yaori

@Pashoke1

ai right now feels powerful but also fragile most models are black boxes you send input you hope the output is honest ⊹ opengradient is attacking that weak point not by building another chatbot but by making model execution verifiable <= what that means => when a model runs you can prove ˟ what version was used ˟ what environment executed it ˟ what output was actually produced that changes trust because today ai outputs are easy to fake easy to manipulate hard to verify ⊹ opengradient connects model execution to cryptographic proof so instead of trust me bro you get mathematical guarantees that is not hype that is infrastructure for serious ai usage finance governance automation anywhere trust matters if ai is going to move real value it needs verifiable execution @OpenGradient is building that layer quietly important

Tech capital

@tfanalytics1

$path vs automation anywhere. It is clear to me that $path does have the best product in workflow automation. If you need simple automations, then you can use power automate from microsoft. If you need really good complex automations that scale, $path is your choice. There isn’t a better automation platform for large enterprises than $path. If you go and see who some of the customers of $path are, most of them are very large enterprises in retail, manufacturing, healthcare - basically the Walmarts, United Healthcare, JP Morgans of the world - household names. Automation anywhere and most other automation tools have small customers that no one has heard of. I envision $path taking the entire automation market for large enterprises because this is where their product scales best and is the best one, while the rest of the market is going to be divided by smaller players.