Think of it as Yoga for the brain, we are going to stretch a bit.
The Panic in Software:
The Monster Is AI —
And It’s Eating Everything
“When AI can do the work, software becomes an expense — and expenses get cut and Wall Street has figured out these companies are becoming obsolete very quickly” --YNOT!
Wall Street smells blood. AI isn’t competing with SaaS. It’s digesting it.
If your product can be described in words, AI can replace it. In every tech revolution, the middle layer gets eaten first.
Nearly $1 trillion in software market value has evaporated in weeks.
The iShares Expanded Tech-Software Sector ETF is down hard. Big names that once traded like royalty are now trading like suspects.
Salesforce. Intuit. Workday. DocuSign.
This isn’t a garden-variety correction. This is existential panic. And the monster causing it?
AI.
The Existential Fear
When a major bank uses the word “existential” about an entire industry, pay attention.
The core fear is simple: If AI agents can perform the task… why pay software subscriptions to do it?
If a workflow can be described in words, an AI model can increasingly execute it.
Draft contracts
File taxes
Reconcile books
Automate HR onboarding
Manage marketing campaigns
Run project dashboards
That’s not incremental improvement. That’s structural compression.
Wall Street sees “seat compression.” You should see revenue compression.
If a CRM charges per user… and AI replaces three users… revenue drops.
If bookkeeping software charges per license… and AI does it automatically… churn rises.
The monster isn’t attacking software margins.
It’s attacking software pricing models.
The Trillion-Dollar Repricing
A year ago, software traded at ~51× earnings.
Today? Roughly half that.
When multiples compress that fast, it’s not about earnings. It’s about revaluation of the entire business model. Software went from: “Scalable recurring revenue” to “Automatable recurring expense.” That’s a dangerous transition.
The Goldman Line in the Sand
But here’s where it gets interesting. Not all software is doomed.
Some of it becomes more valuable in an AI world. The dividing line isn’t hype.
It’s structure.
The Survivors Have At Least One of These:
1️⃣ Physical Infrastructure
If the business owns real computing power, AI cannot eliminate it.
Example:
Microsoft (Azure data centers)
Oracle (AI infrastructure backlog)
Alphabet Inc. (Google Cloud + chips)
AI doesn’t float in space. It runs on servers. And servers cost money.
2️⃣ Regulatory Entrenchment
Banks. Governments. Defense. Healthcare. They don’t switch systems lightly.
Compliance software embedded in regulated environments has inertia.
Example:
Palantir Technologies (defense + government contracts)
The Pentagon is not replacing classified systems with a chatbot.
Switching risk > AI savings.
3️⃣ Deep Operational Integration
If ripping the system out would break the business, the moat remains.
Legacy integrations matter. Especially in large enterprises.
4️⃣ AI Beneficiaries
Some industries get supercharged, not disrupted.
Cybersecurity, for example.
More AI = more attack surface.
Example:
CrowdStrike
Palo Alto Networks
AI creates threats. Security sells defense.
The Vulnerable Class
Now flip the framework.
Software at risk typically has:
Low switching costs
Workflow automation as the core value
Per-user pricing models
Commodity features
Examples often cited:
Salesforce (per-seat CRM exposure)
Intuit (tax + bookkeeping automation risk)
Workday (HR workflow automation)
DocuSign (digital signatures commoditized)
If the product can be reduced to:
“AI can do that.”
You have structural risk.
The Psychological Trap
Retail investors rarely sell winners. They almost never sell losers.
When a stock drops 30%, people “wait to get back to even.”
But markets don’t reward emotional attachment.
They reward structural positioning.
AI is not cyclical.
It is deflationary pressure on software margins.
The question is not: “Will this stock bounce?”
The question is: “Is this business model structurally protected?”
The Bigger Rotation
Here’s the twist.
Some sectors may benefit more from AI than software does.
Transportation and logistics, for example.
AI route optimization.
Predictive maintenance.
Inventory efficiency.
Reduced empty miles.
In some cases, projected profit growth in these sectors dwarfs traditional SaaS growth.
When AI makes physical industries more efficient, margins expand.
The irony?
The “boring” sectors may quietly outperform the former darlings.
The Framework: How to Think About It
Instead of reacting emotionally to red charts, apply this:
