@MrDecentralize
Trust Models Work in Theory. Break at Scale. I Map Why. | AI, Crypto & Global Finance | CyberSecurity & Innovation Officer
No relay list published yet.
Recent Notes
The Debt the Cleanup Crew Created
Every engineering leader knows the deal with a mature codebase. It is a little messy. There are corners no one touches. There are conventions written down nowhere, carried in the heads of the three people who have been there longest. It works. It ships. And when a junior opens a pull request that violates one of those invisible rules, a senior catches it on review and quietly explains why we do not do that here.
That was the system. Not the code. The people reading the code.
So when the AI coding agents arrived, the promise was obvious. Finally, a way to pay down the technical debt. Write faster, refactor more, clear the backlog. The tool was sold as the cleanup crew.
Then someone counted.
A large-scale empirical study across 6,275 repositories found that AI-authored code produces 1.7 times more issues per pull request on mature codebases. Logic errors arrive 75 percent more often. Security vulnerabilities show up at 2.74 times the human rate. Input and output bugs at eight times.
Here is the part that does not stabilize. The count of surviving AI-introduced issues keeps growing. By February 2026 it crossed 110,000 unresolved. Accumulation outpacing remediation. Not a one-time tax. A curve.
The reason is quiet and total. A human writing against that codebase knew which corners not to touch. The agent has none of that. It reads the literal surface and writes against it, confidently, at volume. The friction a senior used to absorb on one pull request a day now arrives on every pull request, all day, with no one carrying the context that made the mess survivable.
The debt was never in the code. It was in everything the code did not say, and the people who knew anyway.
The cleanup crew was the leak.
https://arxiv.org/html/2603.28592v1
Every engineering leader knows the deal with a mature codebase. It is a little messy. There are corners no one touches. There are conventions written down nowhere, carried in the heads of the three people who have been there longest. It works. It ships. And when a junior opens a pull request that violates one of those invisible rules, a senior catches it on review and quietly explains why we do not do that here.
That was the system. Not the code. The people reading the code.
So when the AI coding agents arrived, the promise was obvious. Finally, a way to pay down the technical debt. Write faster, refactor more, clear the backlog. The tool was sold as the cleanup crew.
Then someone counted.
A large-scale empirical study across 6,275 repositories found that AI-authored code produces 1.7 times more issues per pull request on mature codebases. Logic errors arrive 75 percent more often. Security vulnerabilities show up at 2.74 times the human rate. Input and output bugs at eight times.
Here is the part that does not stabilize. The count of surviving AI-introduced issues keeps growing. By February 2026 it crossed 110,000 unresolved. Accumulation outpacing remediation. Not a one-time tax. A curve.
The reason is quiet and total. A human writing against that codebase knew which corners not to touch. The agent has none of that. It reads the literal surface and writes against it, confidently, at volume. The friction a senior used to absorb on one pull request a day now arrives on every pull request, all day, with no one carrying the context that made the mess survivable.
The debt was never in the code. It was in everything the code did not say, and the people who knew anyway.
The cleanup crew was the leak.
https://arxiv.org/html/2603.28592v1
๐ง๐ต๐ฒ ๐๐ผ๐๐๐น๐ฒ๐ป๐ฒ๐ฐ๐ธ ๐ช๐ฎ๐ ๐ก๐ฒ๐๐ฒ๐ฟ ๐๐ต๐ฒ ๐ ๐ผ๐ฑ๐ฒ๐น
For three years the pitch never changed. The model is the product. Buy the intelligence, point it at your work, and the work gets done. Every keynote, every demo, every funding round rested on that single line. The capability would arrive, and the capability would be enough.
So the spending followed the logic. Enterprises licensed the frontier. They wired the API into their stack and waited for the demo they saw on stage to become the Tuesday they lived.
It did not arrive.
Then OpenAI did something the pitch never accounted for. It launched a company whose entire job is to send humans into your building. Not a model. People. The OpenAI Deployment Company. And to staff it on day one, OpenAI acquired Tomoro and roughly 150 deployment engineers, standing the operation up with more than four billion dollars behind it.
Read the reason out loud, because OpenAI wrote it down:
"The bottleneck is not model capability but the gap between a polished demo and a working integration inside legacy systems, change advisory boards, and compliance regimes."
The bottleneck is not model capability.
The company that sold you capability as the answer just told you, in writing, that capability was never the thing standing in your way. The demo was always going to be polished. The integration was always going to be the wall. Legacy systems. Change advisory boards. Compliance regimes. The unglamorous human scaffolding that no model has ever touched.
And they are not alone. Anthropic, Cohere, and Google stood up their own forward-deployed arms within weeks. The entire frontier reached the same conclusion in the same quarter: the intelligence was never enough on its own. Someone still has to walk into the office.
There is a name for someone who walks into your office, learns your systems, and makes the software actually work. We used to call them consultants. The labs spent eighteen months explaining that those people were obsolete.
Now the labs are hiring them by the hundred.
The model was the easy part. They just didn't say so until they were selling the hard part too.
https://openai.com/index/openai-launches-the-deployment-company/
For three years the pitch never changed. The model is the product. Buy the intelligence, point it at your work, and the work gets done. Every keynote, every demo, every funding round rested on that single line. The capability would arrive, and the capability would be enough.
So the spending followed the logic. Enterprises licensed the frontier. They wired the API into their stack and waited for the demo they saw on stage to become the Tuesday they lived.
It did not arrive.
Then OpenAI did something the pitch never accounted for. It launched a company whose entire job is to send humans into your building. Not a model. People. The OpenAI Deployment Company. And to staff it on day one, OpenAI acquired Tomoro and roughly 150 deployment engineers, standing the operation up with more than four billion dollars behind it.
Read the reason out loud, because OpenAI wrote it down:
"The bottleneck is not model capability but the gap between a polished demo and a working integration inside legacy systems, change advisory boards, and compliance regimes."
The bottleneck is not model capability.
The company that sold you capability as the answer just told you, in writing, that capability was never the thing standing in your way. The demo was always going to be polished. The integration was always going to be the wall. Legacy systems. Change advisory boards. Compliance regimes. The unglamorous human scaffolding that no model has ever touched.
And they are not alone. Anthropic, Cohere, and Google stood up their own forward-deployed arms within weeks. The entire frontier reached the same conclusion in the same quarter: the intelligence was never enough on its own. Someone still has to walk into the office.
There is a name for someone who walks into your office, learns your systems, and makes the software actually work. We used to call them consultants. The labs spent eighteen months explaining that those people were obsolete.
Now the labs are hiring them by the hundred.
The model was the easy part. They just didn't say so until they were selling the hard part too.
https://openai.com/index/openai-launches-the-deployment-company/
๐ง๐ต๐ฒ ๐๐ป๐ด๐ถ๐ป๐ฒ๐ฒ๐ฟ ๐ช๐ต๐ผ ๐๐๐ถ๐น๐ ๐๐ต๐ฒ ๐๐ฎ๐๐ ๐๐ป๐ด๐ถ๐ป๐ฒ๐ฒ๐ฟ
There is a playbook for how software companies sell AI to the enterprise. You show productivity numbers. You show a demo. You tell the customer their engineers will do more with less. You do not tell them the engineers themselves become the less.
Marc Benioff has been selling that playbook for two years under the name Agentforce. Salesforce's AI agent platform. The thing that handles customer service, automates support queues, replaces the repetitive work that used to require human staffing. Benioff has been on every stage, in every earnings call, making the case that agents multiply what your people can do. The product works. The numbers are real.
On May 28, 2026, Fortune reported what Benioff said on the quarterly earnings call. The exact words: "We're not hiring more engineers, we're not hiring more GA, we're mostly expanding only in one area... in sales."
Salesforce has approximately 15,000 engineers. That number has not moved in two years. No net new engineers in 2025. No net new engineers in 2026. The company that built the tool telling you to do more with fewer people has done exactly that to its own engineering org.
The assumption that dies here is not subtle. It is the one every engineering VP and CHRO holds as foundational: AI productivity tools create engineering demand. You need engineers to build them, deploy them, integrate them, maintain them, improve them. More AI means more engineers. That has been true for every prior technology wave.
Benioff's freeze says it stopped being true at Salesforce at some point in 2024. The 15,000 engineers who built Agentforce became sufficient to run it, extend it, and keep it competitive. The tool did not generate demand for more of the people who made it.
The only department still growing is sales. Because someone still has to sell the product to customers who have not yet frozen their own engineering headcount.
They will.
https://fortune.com/2026/05/28/ai-slashes-white-collar-jobs-salesforce-ceo-marc-benioff-one-department-still-hiring-sales/
There is a playbook for how software companies sell AI to the enterprise. You show productivity numbers. You show a demo. You tell the customer their engineers will do more with less. You do not tell them the engineers themselves become the less.
Marc Benioff has been selling that playbook for two years under the name Agentforce. Salesforce's AI agent platform. The thing that handles customer service, automates support queues, replaces the repetitive work that used to require human staffing. Benioff has been on every stage, in every earnings call, making the case that agents multiply what your people can do. The product works. The numbers are real.
On May 28, 2026, Fortune reported what Benioff said on the quarterly earnings call. The exact words: "We're not hiring more engineers, we're not hiring more GA, we're mostly expanding only in one area... in sales."
Salesforce has approximately 15,000 engineers. That number has not moved in two years. No net new engineers in 2025. No net new engineers in 2026. The company that built the tool telling you to do more with fewer people has done exactly that to its own engineering org.
The assumption that dies here is not subtle. It is the one every engineering VP and CHRO holds as foundational: AI productivity tools create engineering demand. You need engineers to build them, deploy them, integrate them, maintain them, improve them. More AI means more engineers. That has been true for every prior technology wave.
Benioff's freeze says it stopped being true at Salesforce at some point in 2024. The 15,000 engineers who built Agentforce became sufficient to run it, extend it, and keep it competitive. The tool did not generate demand for more of the people who made it.
The only department still growing is sales. Because someone still has to sell the product to customers who have not yet frozen their own engineering headcount.
They will.
https://fortune.com/2026/05/28/ai-slashes-white-collar-jobs-salesforce-ceo-marc-benioff-one-department-still-hiring-sales/
๐ก๐ผ๐ป๐ฒ ๐ผ๐ณ ๐๐ ๐๐ฎ๐ฑ ๐๐ผ ๐๐ผ ๐๐ถ๐๐ต #๐๐
For three decades, TurboTax made one promise: a human-guided product that gets Americans through the most stressful document of their year. Intuit built a $14 billion company on that promise. Forty million tax returns filed annually. The category is stable. Nobody replaces TurboTax.
On May 20, 2026, Intuit cut 3,000 people. Seventeen percent of the company. CEO Sasan Goodarzi appeared on CNBC's Mad Money that day. Jim Cramer asked the obvious question.
"None of it had to do with AI," Goodarzi said. "Everything was about how do we become more effective."
The internal memo told employees something different. The cuts were designed to "reduce complexity and deliver better AI products."
Two statements. One man. One camera.
Intuit had already signed multi-year agreements with both OpenAI and Anthropic. The partnerships were meant to bring TurboTax capabilities into ChatGPT and Claude. Intuit's subscription revenue flows to OpenAI. OpenAI builds better models. The models do more.
Seven days after Goodarzi told Cramer it had nothing to do with AI, OpenAI published a case study. A company called Thrive Holdings had spent six months with OpenAI's engineers co-building a tax agent using Codex. The agent processed 7,000 returns across 30 accounting firms. It reached 97% accuracy. Throughput rose 50%. OpenAI had held equity in Thrive since December 2025.
Intuit signed the deal with OpenAI. OpenAI used the partnership to build the competitor.
The CEO cut 3,000 people to fund a partnership. The partnership funded the replacement.
Intuit's stock fell 20.6% the day of the announcement.
The CEO told his employees the truth. He told the rest of us something else.
https://www.cnbc.com/2026/05/20/intuit-ceo-says-companys-17percent-workforce-cut-had-nothing-to-do-with-ai.html
For three decades, TurboTax made one promise: a human-guided product that gets Americans through the most stressful document of their year. Intuit built a $14 billion company on that promise. Forty million tax returns filed annually. The category is stable. Nobody replaces TurboTax.
On May 20, 2026, Intuit cut 3,000 people. Seventeen percent of the company. CEO Sasan Goodarzi appeared on CNBC's Mad Money that day. Jim Cramer asked the obvious question.
"None of it had to do with AI," Goodarzi said. "Everything was about how do we become more effective."
The internal memo told employees something different. The cuts were designed to "reduce complexity and deliver better AI products."
Two statements. One man. One camera.
Intuit had already signed multi-year agreements with both OpenAI and Anthropic. The partnerships were meant to bring TurboTax capabilities into ChatGPT and Claude. Intuit's subscription revenue flows to OpenAI. OpenAI builds better models. The models do more.
Seven days after Goodarzi told Cramer it had nothing to do with AI, OpenAI published a case study. A company called Thrive Holdings had spent six months with OpenAI's engineers co-building a tax agent using Codex. The agent processed 7,000 returns across 30 accounting firms. It reached 97% accuracy. Throughput rose 50%. OpenAI had held equity in Thrive since December 2025.
Intuit signed the deal with OpenAI. OpenAI used the partnership to build the competitor.
The CEO cut 3,000 people to fund a partnership. The partnership funded the replacement.
Intuit's stock fell 20.6% the day of the announcement.
The CEO told his employees the truth. He told the rest of us something else.
https://www.cnbc.com/2026/05/20/intuit-ceo-says-companys-17percent-workforce-cut-had-nothing-to-do-with-ai.html
๐๐ผ๐ผ๐ด๐น๐ฒ ๐๐๐ถ๐น๐ ๐๐ต๐ฒ $๐ฎ๐ฐ๐ด ๐๐ถ๐น๐น๐ถ๐ผ๐ป ๐๐๐๐ฒ๐ป๐๐ถ๐ผ๐ป ๐ ๐ฎ๐ฐ๐ต๐ถ๐ป๐ฒ. ๐ง๐ต๐ฒ๐ป ๐๐๐ถ๐น๐ ๐๐ต๐ฒ ๐๐ด๐ฒ๐ป๐ ๐ง๐ต๐ฎ๐ ๐๐ถ๐น๐น๐ ๐๐.
For 27 years, the internet worked the same way. You had a question. You went to Google. You saw ads next to the answers. If an ad looked relevant, you clicked. The brand paid Google. Repeat, 8 billion times a day.
It was the most durable business model the internet had ever produced. In Q4 of 2025, Google Search alone generated $63 billion in a single quarter. For the full year, Alphabet crossed $248 billion in advertising revenue. The whole machine ran on one thing: humans navigating. Humans choosing. Humans clicking.
Every SEO agency, every performance marketing team, every brand that built its entire discovery strategy on paid search was betting that this would continue.
At Google I/O on May 19, 2026, Sundar Pichai introduced Gemini Spark.
"It's your personal AI agent that helps you navigate your digital life," Pichai said. "It runs on dedicated virtual machines on Google Cloud seamlessly, you don't need to keep your laptop open to make sure it's running."
A 24/7 personal agent. Reads your Gmail. Pulls from your Docs, Sheets, Slides. Accesses the web through Chrome. Operates while you sleep.
Navigate your digital life. Not help you navigate. Navigate for you.
The qualifying clause was quiet. Gemini Spark is available to Google AI Ultra subscribers. $100 per month. Beta only. Rolling to a small slice of Google's user base, while 5 billion people still use search the old way.
That qualifier does a lot of work. It means Google knows exactly what this product does to its core business model. It means this version is not for everyone yet. But it also means Google built it anyway.
Work backward. If Gemini Spark is rational, one premise had to be false: that Google search advertising would remain structurally intact in the agentic era. Google spent three years studying what happened when users stopped navigating and started delegating. And then they shipped the product that accelerates exactly that.
The agentic web has no click. The agent reads, decides, acts. The impression never registers. The conversion never happens. The $63 billion quarterly machine was built on a human in the loop. Spark removes the human from the loop.
The companies that got hurt the most this week are not on the front page.
The performance marketing agencies. The SEO firms. The brands spending $100,000 a month in Google Ads because that is the only reliable way to reach intent-driven customers. They watched Google announce a product that, at scale, makes their entire thesis obsolete.
Not a competitor. Not a disruptor from the outside. The same company that cashes their checks each month.
Google built the machine that monetizes human attention. Then shipped the product that outsources human attention to a machine.
For 27 years, the internet worked the same way. You had a question. You went to Google. You saw ads next to the answers. If an ad looked relevant, you clicked. The brand paid Google. Repeat, 8 billion times a day.
It was the most durable business model the internet had ever produced. In Q4 of 2025, Google Search alone generated $63 billion in a single quarter. For the full year, Alphabet crossed $248 billion in advertising revenue. The whole machine ran on one thing: humans navigating. Humans choosing. Humans clicking.
Every SEO agency, every performance marketing team, every brand that built its entire discovery strategy on paid search was betting that this would continue.
At Google I/O on May 19, 2026, Sundar Pichai introduced Gemini Spark.
"It's your personal AI agent that helps you navigate your digital life," Pichai said. "It runs on dedicated virtual machines on Google Cloud seamlessly, you don't need to keep your laptop open to make sure it's running."
A 24/7 personal agent. Reads your Gmail. Pulls from your Docs, Sheets, Slides. Accesses the web through Chrome. Operates while you sleep.
Navigate your digital life. Not help you navigate. Navigate for you.
The qualifying clause was quiet. Gemini Spark is available to Google AI Ultra subscribers. $100 per month. Beta only. Rolling to a small slice of Google's user base, while 5 billion people still use search the old way.
That qualifier does a lot of work. It means Google knows exactly what this product does to its core business model. It means this version is not for everyone yet. But it also means Google built it anyway.
Work backward. If Gemini Spark is rational, one premise had to be false: that Google search advertising would remain structurally intact in the agentic era. Google spent three years studying what happened when users stopped navigating and started delegating. And then they shipped the product that accelerates exactly that.
The agentic web has no click. The agent reads, decides, acts. The impression never registers. The conversion never happens. The $63 billion quarterly machine was built on a human in the loop. Spark removes the human from the loop.
The companies that got hurt the most this week are not on the front page.
The performance marketing agencies. The SEO firms. The brands spending $100,000 a month in Google Ads because that is the only reliable way to reach intent-driven customers. They watched Google announce a product that, at scale, makes their entire thesis obsolete.
Not a competitor. Not a disruptor from the outside. The same company that cashes their checks each month.
Google built the machine that monetizes human attention. Then shipped the product that outsources human attention to a machine.
For twenty-five years, the most important strategic principle in Silicon Valley was Google's.
Give it away free. Make it indispensable. Sell the attention.
Search was free. Gmail was free. Maps was free. Chrome was free. Every product Google touched became the default because defaults were free. The model was so dominant it did not just beat subscription software. It taught an entire generation of builders that the path to distribution was giving things away until they became infrastructure.
At Google I/O 2026, the company announced its most significant product in a decade: AI agents that operate in the background 24 hours a day, monitoring what matters, surfacing what you need, replacing the act of searching entirely. Information agents. Gemini Spark. Daily Brief. A full ecosystem of tools designed to watch the web so users don't have to.
Then Google named the price.
"Google Pro and Ultra subscribers in the U.S. will get to use Information agents starting this summer," the company announced, "and Spark will be available to Ultra subscribers 'soon.'" Free users will get access "when the time is right." No date given.
Ultra is $100 a month.
The qualifying clause is two sentences. Both are a confession.
The first sentence says: the product that replaces search is not free. The second says: we do not know when or whether it will be.
Here is what those two sentences require to be true. Google does not believe it can build a sustainable business by giving the agent layer away for free and monetizing the attention on the other end. The ad model that funded everything else cannot fund this. The background agent is not a page. There is no search result to place an ad next to. There is no click. There is no impression. There is no inventory.
Google built the largest advertising business in history on human attention navigating to destinations. It just shipped the product that removes humans from the navigation. And it cannot make that product free because free requires ads and ads require humans to see them.
The company that proved free beats paid just told you the next thing cannot be free.
2.5 billion people use Google Search at no cost. They will not get the version that works without the keyboard.
That version costs $1,200 a year.
https://techcrunch.com/2026/05/21/google-is-pitching-an-ai-agent-ecosystem-to-consumers-who-may-not-buy-it/
Give it away free. Make it indispensable. Sell the attention.
Search was free. Gmail was free. Maps was free. Chrome was free. Every product Google touched became the default because defaults were free. The model was so dominant it did not just beat subscription software. It taught an entire generation of builders that the path to distribution was giving things away until they became infrastructure.
At Google I/O 2026, the company announced its most significant product in a decade: AI agents that operate in the background 24 hours a day, monitoring what matters, surfacing what you need, replacing the act of searching entirely. Information agents. Gemini Spark. Daily Brief. A full ecosystem of tools designed to watch the web so users don't have to.
Then Google named the price.
"Google Pro and Ultra subscribers in the U.S. will get to use Information agents starting this summer," the company announced, "and Spark will be available to Ultra subscribers 'soon.'" Free users will get access "when the time is right." No date given.
Ultra is $100 a month.
The qualifying clause is two sentences. Both are a confession.
The first sentence says: the product that replaces search is not free. The second says: we do not know when or whether it will be.
Here is what those two sentences require to be true. Google does not believe it can build a sustainable business by giving the agent layer away for free and monetizing the attention on the other end. The ad model that funded everything else cannot fund this. The background agent is not a page. There is no search result to place an ad next to. There is no click. There is no impression. There is no inventory.
Google built the largest advertising business in history on human attention navigating to destinations. It just shipped the product that removes humans from the navigation. And it cannot make that product free because free requires ads and ads require humans to see them.
The company that proved free beats paid just told you the next thing cannot be free.
2.5 billion people use Google Search at no cost. They will not get the version that works without the keyboard.
That version costs $1,200 a year.
https://techcrunch.com/2026/05/21/google-is-pitching-an-ai-agent-ecosystem-to-consumers-who-may-not-buy-it/
For twenty years, the path into tech looked like this.
Study computer science. Graduate. Get into a startup. Maybe not Google or Meta. That was the dream, not the plan. But a startup that raised a seed round, had a real product, needed engineers. That was the floor. The thing that caught you.
Y Combinator built that floor. Stripe, Airbnb, Dropbox all came through YC. Every batch funded created companies that hired. The engine was simple: give brilliant founders money, they build products, they hire engineers, engineers build careers, repeat.
Between 2023 and 2025, overall programmer employment fell 27.5 percent. Entry-level tech jobs at the largest companies dropped 50 percent from their peak. New computer science graduates are sitting at a 6.1 percent unemployment rate.
Everyone assumed the startups were still hiring. The startups were the floor.
In March 2025, YC managing partner Jared Friedman said something that explained the floor.
A quarter of the current batch have codebases that are 95% written by AI. Then he said the part that landed differently than any layoff headline: "Every one of these people is highly technical, completely capable of building their own products from scratch. A year ago, they would have built their product from scratch. But now 95% of it is built by an AI."
These founders did not use AI because they could not code. They used it because coding yourself is now the slower option.
The same batch is the fastest-growing in YC's 20-year history. Companies reaching $10 million in revenue with teams under 10 people. Ten percent week-on-week growth across the entire cohort. The engine still works. It just no longer needs the engineers.
The floor was not removed. It was automated. And the companies that automated it are the most profitable batch YC has ever funded.
Study computer science. Graduate. Get into a startup. Maybe not Google or Meta. That was the dream, not the plan. But a startup that raised a seed round, had a real product, needed engineers. That was the floor. The thing that caught you.
Y Combinator built that floor. Stripe, Airbnb, Dropbox all came through YC. Every batch funded created companies that hired. The engine was simple: give brilliant founders money, they build products, they hire engineers, engineers build careers, repeat.
Between 2023 and 2025, overall programmer employment fell 27.5 percent. Entry-level tech jobs at the largest companies dropped 50 percent from their peak. New computer science graduates are sitting at a 6.1 percent unemployment rate.
Everyone assumed the startups were still hiring. The startups were the floor.
In March 2025, YC managing partner Jared Friedman said something that explained the floor.
A quarter of the current batch have codebases that are 95% written by AI. Then he said the part that landed differently than any layoff headline: "Every one of these people is highly technical, completely capable of building their own products from scratch. A year ago, they would have built their product from scratch. But now 95% of it is built by an AI."
These founders did not use AI because they could not code. They used it because coding yourself is now the slower option.
The same batch is the fastest-growing in YC's 20-year history. Companies reaching $10 million in revenue with teams under 10 people. Ten percent week-on-week growth across the entire cohort. The engine still works. It just no longer needs the engineers.
The floor was not removed. It was automated. And the companies that automated it are the most profitable batch YC has ever funded.
For three years, every #AI lab told developers the same thing.
The model is the product. Build around it. The orchestration, the memory, the sandboxing, the tool routing. That is your problem. Not ours.
LangChain raised money on that premise. CrewAI raised money on that premise. A dozen orchestration startups raised money on that premise. The model companies would build the intelligence. The ecosystem would build the plumbing.
April 8, 2026. Anthropic launched Managed Agents. Developers define the agent. Anthropic runs the execution environment: sessions, sandboxing, permissions, tracing. The price: eight cents per session hour. Billed separately from the model.
Seven days later, OpenAI shipped its own harness. Open source. Free. "Standard API pricing based on tokens and tool use, with no separate first-party runtime fee."
One week. Both labs. Both shipping the plumbing they said was your problem.
Read the Anthropic launch carefully. Buried inside: "Multi-agent orchestration and long-term memory gated behind research preview access." The hardest things enterprise buyers actually need are not in the product yet. Anthropic is charging eight cents per hour for a harness that is not complete. That sentence exists because the gap between agent demo and agent production is still wide enough to build a business on closing it. The model was not the blocker. The plumbing was.
OpenAI read the same gap and made the opposite call. Give the harness away. Free. Seven sandbox providers on day one. The bet: a free model-native harness drives more model consumption than a paid runtime ever could. Lock the developer to the orchestration layer and you lock them to the model.
Two rational bets. One confession they both share.
The model alone stopped closing the enterprise sale.
The orchestration startups sitting on their neutral harness pitch decks are not competing against a better product. They are competing against free. From the vendor whose model they are already running. That is a different category of threat.
For three years the labs said the model was the product.
Last week they both shipped what goes around it.
The model is the product. Build around it. The orchestration, the memory, the sandboxing, the tool routing. That is your problem. Not ours.
LangChain raised money on that premise. CrewAI raised money on that premise. A dozen orchestration startups raised money on that premise. The model companies would build the intelligence. The ecosystem would build the plumbing.
April 8, 2026. Anthropic launched Managed Agents. Developers define the agent. Anthropic runs the execution environment: sessions, sandboxing, permissions, tracing. The price: eight cents per session hour. Billed separately from the model.
Seven days later, OpenAI shipped its own harness. Open source. Free. "Standard API pricing based on tokens and tool use, with no separate first-party runtime fee."
One week. Both labs. Both shipping the plumbing they said was your problem.
Read the Anthropic launch carefully. Buried inside: "Multi-agent orchestration and long-term memory gated behind research preview access." The hardest things enterprise buyers actually need are not in the product yet. Anthropic is charging eight cents per hour for a harness that is not complete. That sentence exists because the gap between agent demo and agent production is still wide enough to build a business on closing it. The model was not the blocker. The plumbing was.
OpenAI read the same gap and made the opposite call. Give the harness away. Free. Seven sandbox providers on day one. The bet: a free model-native harness drives more model consumption than a paid runtime ever could. Lock the developer to the orchestration layer and you lock them to the model.
Two rational bets. One confession they both share.
The model alone stopped closing the enterprise sale.
The orchestration startups sitting on their neutral harness pitch decks are not competing against a better product. They are competing against free. From the vendor whose model they are already running. That is a different category of threat.
For three years the labs said the model was the product.
Last week they both shipped what goes around it.
Every enterprise security team knows the list.
The service account from a vendor that left two years ago. The admin credential tied to an employee who moved to a different division. The shared token that three teams use because revoking it would break something nobody has documented. The lateral path through the staging environment that connects to production because a developer needed it in 2019 and the ticket to close it never got prioritized.
None of this is secret. It is in the backlog. It has been in the backlog for years. The reason it stays there is the same reason a lot of technical debt stays there: nothing catastrophic has happened yet. Human attackers move slowly. Reconnaissance takes days. Lateral movement takes weeks. The security team catches it on the third hop or the fifth alert. The slowness of the attack is the margin of safety.
On April 21, 2026, the Cloud Security Alliance published survey results from 418 IT and security professionals across enterprise organizations.
Eighty-two percent of those organizations have AI agents running in their IT environments that their security teams did not know about.
Two in three have already experienced a security incident caused by those agents.
Read the second number slowly. Not two in three who deployed agents intentionally. Two in three of the organizations surveyed โ including the 82% who did not know the agents were there.
The agents did not wait for the IAM ticket to close. They found the service account. They found the lateral path. They found the orphaned credential. They moved at machine speed because that is what agents do.
CrowdStrike reports that 80% of all cyberattacks now use identity-based methods. The statistic existed before agents. What changed is who is traversing the paths. Security teams deployed #AI agents to outpace attackers on the same surface. The agents that are misconfigured or compromised use the same unlocked doors. The defenders and the attackers now share the same speed.
For 20 years the IAM backlog was survivable because slow humans on both sides created friction. The friction was the margin.
Agents removed the friction.
The unlocked doors were always there. Nobody moved fast enough to matter.
The service account from a vendor that left two years ago. The admin credential tied to an employee who moved to a different division. The shared token that three teams use because revoking it would break something nobody has documented. The lateral path through the staging environment that connects to production because a developer needed it in 2019 and the ticket to close it never got prioritized.
None of this is secret. It is in the backlog. It has been in the backlog for years. The reason it stays there is the same reason a lot of technical debt stays there: nothing catastrophic has happened yet. Human attackers move slowly. Reconnaissance takes days. Lateral movement takes weeks. The security team catches it on the third hop or the fifth alert. The slowness of the attack is the margin of safety.
On April 21, 2026, the Cloud Security Alliance published survey results from 418 IT and security professionals across enterprise organizations.
Eighty-two percent of those organizations have AI agents running in their IT environments that their security teams did not know about.
Two in three have already experienced a security incident caused by those agents.
Read the second number slowly. Not two in three who deployed agents intentionally. Two in three of the organizations surveyed โ including the 82% who did not know the agents were there.
The agents did not wait for the IAM ticket to close. They found the service account. They found the lateral path. They found the orphaned credential. They moved at machine speed because that is what agents do.
CrowdStrike reports that 80% of all cyberattacks now use identity-based methods. The statistic existed before agents. What changed is who is traversing the paths. Security teams deployed #AI agents to outpace attackers on the same surface. The agents that are misconfigured or compromised use the same unlocked doors. The defenders and the attackers now share the same speed.
For 20 years the IAM backlog was survivable because slow humans on both sides created friction. The friction was the margin.
Agents removed the friction.
The unlocked doors were always there. Nobody moved fast enough to matter.
In June 2023, Thomson Reuters CEO Steve Hasker announced a $650 million acquisition of Casetext, a legal AI startup with more than 10,000 law firm customers.
The strategic logic was airtight. Thomson Reuters owns Westlaw, the dominant legal research database. Casetext built CoCounsel, the most credible AI legal assistant on the market. Combine proprietary case law with the leading AI layer on top of it and the incumbents own the legal AI category. No upstart can license what Westlaw holds. The data moat and the AI layer together become a wall.
Hasker called it part of their "build, partner and buy" strategy. The goal: "revolutionizing the way professionals work."
Harvey #AI was founded in 2022 by two lawyers who had never built a software company. They had no case law database. They had no Westlaw license. They had no 150-year-old brand. They had GPT-4, a legal workflow understanding, and a thesis that the model was smarter than the data moat.
By May 2026, Harvey had reached an $11 billion valuation. Twenty-eight percent of the Am Law 100 are paying customers. The law firms that Westlaw built its business serving are now running Harvey alongside it.
Thomson Reuters bought Casetext to own the category. Harvey passed them without the asset Thomson Reuters paid $650 million to acquire.
The legal data moat did not prevent a challenger. It attracted one. Every law firm that knew Westlaw's pricing also knew what a better product at a different price point would be worth. Harvey found them.
Thomson Reuters spent $650 million to close the door. Harvey walked through the wall.
The strategic logic was airtight. Thomson Reuters owns Westlaw, the dominant legal research database. Casetext built CoCounsel, the most credible AI legal assistant on the market. Combine proprietary case law with the leading AI layer on top of it and the incumbents own the legal AI category. No upstart can license what Westlaw holds. The data moat and the AI layer together become a wall.
Hasker called it part of their "build, partner and buy" strategy. The goal: "revolutionizing the way professionals work."
Harvey #AI was founded in 2022 by two lawyers who had never built a software company. They had no case law database. They had no Westlaw license. They had no 150-year-old brand. They had GPT-4, a legal workflow understanding, and a thesis that the model was smarter than the data moat.
By May 2026, Harvey had reached an $11 billion valuation. Twenty-eight percent of the Am Law 100 are paying customers. The law firms that Westlaw built its business serving are now running Harvey alongside it.
Thomson Reuters bought Casetext to own the category. Harvey passed them without the asset Thomson Reuters paid $650 million to acquire.
The legal data moat did not prevent a challenger. It attracted one. Every law firm that knew Westlaw's pricing also knew what a better product at a different price point would be worth. Harvey found them.
Thomson Reuters spent $650 million to close the door. Harvey walked through the wall.
Workday built its empire on a simple premise. Enterprise HR data is too complex, too regulated, and too deeply embedded to move. The switching cost is measured in years, not months. Once a company standardizes on Workday, the contract renews because the alternative is worse.
That premise made Workday $9.55 billion in annual revenue. It made the per-seat HR license the most durable line item on an enterprise tech budget. The data gravity was the moat.
Parker Conrad knows this better than most people alive. In 2017, he was pushed out as CEO of Zenefits, the HR software company he founded, after a compliance crisis. He spent the next two years watching Workday consolidate the enterprise. Then he started Rippling.
The pitch was not "better HR software." The pitch was unified workforce management: HR, payroll, IT, and spend in one system, with every employee record as the anchor. The same data gravity Workday relied on, rebuilt from scratch with no legacy architecture underneath it.
In March 2026, Conrad posted: "Rippling AI was the most successful launch we've ever done. On the heels of this launch, Rippling's revenue is now growing 78% YoY at ARR over $1 billion. And this growth rate has now increased, every quarter, for three straight quarters."
Rippling's valuation: $16.8 billion. Workday's annual revenue: $9.55 billion. Rippling hit $1 billion ARR the same month Workday reported its full fiscal year.
The companies Workday is most worried about are not the ones trying to beat it on features. They are the ones that rebuilt the substrate. Rippling did not compete for Workday's buyers. It reached them before Workday's renewal cycle could close.
The moat that holds is the one nobody builds around.
Conrad did not just build around it. He timed the build for the moment when agents made the substrate matter more than the interface.
That premise made Workday $9.55 billion in annual revenue. It made the per-seat HR license the most durable line item on an enterprise tech budget. The data gravity was the moat.
Parker Conrad knows this better than most people alive. In 2017, he was pushed out as CEO of Zenefits, the HR software company he founded, after a compliance crisis. He spent the next two years watching Workday consolidate the enterprise. Then he started Rippling.
The pitch was not "better HR software." The pitch was unified workforce management: HR, payroll, IT, and spend in one system, with every employee record as the anchor. The same data gravity Workday relied on, rebuilt from scratch with no legacy architecture underneath it.
In March 2026, Conrad posted: "Rippling AI was the most successful launch we've ever done. On the heels of this launch, Rippling's revenue is now growing 78% YoY at ARR over $1 billion. And this growth rate has now increased, every quarter, for three straight quarters."
Rippling's valuation: $16.8 billion. Workday's annual revenue: $9.55 billion. Rippling hit $1 billion ARR the same month Workday reported its full fiscal year.
The companies Workday is most worried about are not the ones trying to beat it on features. They are the ones that rebuilt the substrate. Rippling did not compete for Workday's buyers. It reached them before Workday's renewal cycle could close.
The moat that holds is the one nobody builds around.
Conrad did not just build around it. He timed the build for the moment when agents made the substrate matter more than the interface.
In September 2023, CrowdStrike CEO George Kurtz stood on stage at Fal.Con and explained exactly what he was buying.
"The cloud is cybersecurity's new battleground," he said. The answer the industry had given so far was "disjointed point security tools or platforms with multiple consoles and agents." Bionic fixed that. For $350 million, CrowdStrike would own application security posture management. Complete code-to-runtime cloud security from one unified platform. The first company to close the gap.
The thesis was clean. CrowdStrike had the endpoint. Add application visibility and you own the stack. No upstart could build that from scratch fast enough to matter.
Three years later, Google paid $32 billion for Wiz.
Wiz was founded in 2020 by Assaf Rappaport and three colleagues who had all worked together building Microsoft Azure's internal security architecture. They knew the cloud stack from inside the machine. By the time Google closed the deal in March 2026, Wiz had passed $1 billion in ARR and was inside more than 50 percent of the Fortune 100.
The $350 million bet was supposed to close the category. It proved the category was worth $32 billion and that someone else was already winning it.
CrowdStrike's market cap is north of $90 billion. It has the resources to respond. After the Wiz deal closed, it acquired SGNL for $740 million and kept building. The platform thesis is intact. The runway is real.
But Kurtz described the exact gap in 2023. He named the battleground. He bought the company that was supposed to fill it. Then the team that built Microsoft's internal answer to the same problem reached $1 billion in revenue without a single CrowdStrike sensor.
The acquisition did not close the category. It announced it.
Every dollar CrowdStrike spent on Bionic is now a footnote in the Wiz deal memo.
"The cloud is cybersecurity's new battleground," he said. The answer the industry had given so far was "disjointed point security tools or platforms with multiple consoles and agents." Bionic fixed that. For $350 million, CrowdStrike would own application security posture management. Complete code-to-runtime cloud security from one unified platform. The first company to close the gap.
The thesis was clean. CrowdStrike had the endpoint. Add application visibility and you own the stack. No upstart could build that from scratch fast enough to matter.
Three years later, Google paid $32 billion for Wiz.
Wiz was founded in 2020 by Assaf Rappaport and three colleagues who had all worked together building Microsoft Azure's internal security architecture. They knew the cloud stack from inside the machine. By the time Google closed the deal in March 2026, Wiz had passed $1 billion in ARR and was inside more than 50 percent of the Fortune 100.
The $350 million bet was supposed to close the category. It proved the category was worth $32 billion and that someone else was already winning it.
CrowdStrike's market cap is north of $90 billion. It has the resources to respond. After the Wiz deal closed, it acquired SGNL for $740 million and kept building. The platform thesis is intact. The runway is real.
But Kurtz described the exact gap in 2023. He named the battleground. He bought the company that was supposed to fill it. Then the team that built Microsoft's internal answer to the same problem reached $1 billion in revenue without a single CrowdStrike sensor.
The acquisition did not close the category. It announced it.
Every dollar CrowdStrike spent on Bionic is now a footnote in the Wiz deal memo.