Episode summary
Why AI Won't Transform Most Enterprises for 10 Years
How I Invest With David Weisburd
· · 57 min · Hosts: David Weisburd · Guests: Sushanth Raman
Summary
Sushanth Raman argues that AI will not transform most Fortune 500 companies overnight because enterprise adoption is constrained by organizational change, legacy infrastructure, and undocumented tribal knowledge. He says the real bottleneck is not model quality but whether CEOs, CTOs, and operators are personally committed to rewiring workflows, incentives, and decision rights around the new tools.
He sees the fastest adoption in software engineering and other more deterministic tasks, but believes logistics and broader supply chain operations are especially attractive because they are messy, high-volume, and economically important. In his view, the best AI products will not simply assist employees; they will execute work inside existing enterprise systems, especially where accuracy and context matter more than generic automation.
Raman also draws a sharp line between thin application software and durable systems of record. He thinks some SaaS companies are exposed as software becomes cheaper to build, while platforms with deep data, workflow ownership, and security relevance may remain sticky or even benefit. For Pallet, that means building a context-rich execution layer that improves as foundation models improve, rather than tying the business to any one model release.
Key takeaways
- Sushanth Raman expects most enterprises to need 5 to 10 years before AI is deeply adopted because change management, legacy systems, and tribal knowledge slow deployment.
- He says AI adoption works best when the CEO, CTO, COO, and business leaders are personally involved, not when a head of AI is left isolated.
- Raman believes logistics and supply chain are unusually strong AI markets because they are high-stakes, data-poor, and full of operational nuance.
- He thinks thin application software is most vulnerable in the AI era, while systems of record and security vendors may be more durable or even advantaged.
- He says Pallet is built to get better as foundation models improve by combining context, data integration, and execution layers rather than hard-coded logic.
Notable quotes
“AI won't transform most companies for another 5 to 10 years”
“the number one determinant is when leadership is really bought in top down on like we're going to personally go and drive this change”
“It's not like you can go like there's an AI person, you take care of AI. It's like you yourself have to make this a part of your job”
“this is not something that could be vibe code. This is not something where even 99% accuracy is good enough. It has to be 99.9% accurate”
“Enterprise software companies don't look anything like enterprise software companies of the past”
How the episode opens
Everyone in Silicon Valley talks as if AI is about to transform every company. Today's guest thinks that's wrong. Shush Raman is the founder and CEO of Pallet, an AI company backed by General Catalyst, Bank Capital Ventures, and Bessemer that helps Fortune 500 companies deploy AI into real business operations.
After working with some of the world's largest enterprises, Shush believes AI won't transform most companies for another 5 to 10 years. Not because the technology isn't improving, but because most organizations aren't built to take advantage of it. Without further ado, here's my conversation with Sushan.
Silicon Valley talks as if AI is going to transform every company overnight. You think Fortune 500 companies are 5 to 10 years away. Why?
Silicon Valley has this conception that just because the technology is here, it's really easy to get distributed everywhere in the world. But if you look at these Fortune 500 companies, there's a lot that they need to kind of get right from first of all, these companies have a lot of organizational change that they need to drive, right? Like it's not like as soon as the seuite decides that we should do this, it's going to get done.
You need to kind of figure out how do they reorg their company? How did they change these roles? How do they kind of scope that out?
Secondly, their technology stack is often not ready. They have like a lot of on-prem systems, lots of integrations that are hard to deal with, legacy tech stacks. And then thirdly, there's a lot of like tribal knowledge, poorly documented institutional knowledge.
It's like not well captured. So even if you deploy the technology properly, you're going to find that it's going to be limited in the kind of context it has to kind of go and succeed.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- enterprise ai adoption
- reasoning models
- logistics supply chain
- ai workflow automation
- saas apocalypse misconception
Companies and tickers mentioned
- Pallet
- Bain Capital Ventures
- General Catalyst
- Bessemer Venture Partners
- Activant
Watch the episode
Watch “Why AI Won't Transform Most Enterprises for 10 Years” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from How I Invest With David Weisburd
-
Hunter Walk: Why Venture is Getting Harder, Picking Great Founders, and Investing Against Consensus -
True Anomaly’s Even Rogers on Space Warfare, China’s Military Advantage, and the Defense Tech Boom -
Finding the Next Anthropic, Quantifying Venture Capital, and the Science of Picking Winners -
Why 95% of AI Projects Fail, Anthropic still at $2T?, and Why AI Is Still in the First Inning -
Peter Thiel's 27-Year Partner on the AI Bubble, the Venture Capital Crisis, and the PayPal Mafia -
Why Venture Is Eating Public Markets, Picking Great VCs, and the AI Boom w/Brijesh Jeevarathnam
How I Invest With David Weisburd, this episode and its artwork belong to David Weisburd. Octagon is not affiliated with the show; this page is an independent summary.
