Exponential Operator is a 100% human-written newsletter by Sawyer Middeleer, founder of Revi Systems. If you were forwarded this email, join some of the sharpest AI operators by subscribing here:
Hi all,
If you haven’t heard, services are apparently the new hot thing in AI.
Yesterday OpenAI launched The Deployment Company, a $10B AI venture backed by a battery of institutional partners with a stated goal “to drive broader AI adoption.”
Anthropic is developing a similar offering, following their recent partnership with Accenture also with the goal of helping “enterprises move from AI pilots to full-scale deployment.”
Aaron Levi, CEO of Box sums up the mind of the enterprise customer these services are made for well:
“As agents enter knowledge work beyond coding, there is very real work to upgrade IT systems, get agents the context they need, modernize the workflows to work with agents, figure out the human-agent relationship in the workflow, drive adoption and do change management, and much more.
While AI models have an incredible amount of capability packed into them, there’s no shortcut to getting that intelligence applied to a business process in a stable way. This is creating tons of opportunities across the market for new jobs and firms, and the labs are equally recognizing the criticality here.”
Earlier this week I wrote about the 5 levels of AI adoption. What I should have highlighted better is the large gap I’ve observed for many organizations trying to get from ~L2 to L3 and further.
It’s the first leap in AI adoption where companies start facing significant structural headwinds - IT infrastructure may not support AI agents; internal and external stakeholders might oppose transformation; employees would need months of upskilling; org charts and SOPs aren’t set up for change.
Overcoming any of these challenges individually is a massive undertaking with tons of moving parts. Overhauling a company entirely around AI multiplies the difficulty further. Even with leadership all on the same page, the critical gap that organizations are facing - this is true for mature companies of almost any size, not just enterprises - is complexity.
Yoni Rechtman identified the crisis of complexity born from accelerating AI adoption a year ago:
As software gets better, enterprise environments and IT systems are going to get way more complex and harder to manage.
Software/AI can do more things ▶️ more use cases
Cost to produce code goes down ▶️ more options/vendors for each use case
Products get better faster ▶️ more features/releases to manage
Collectively, that means the future looks like way more apps and systems from more vendors swapped out more frequently with more and smaller vendors to evaluate each time you do. At the same time, MCP, operator models, etc. create entirely new surface areas and vectors for complexity, integrations, and management.
His analysis strikes at something bigger beyond this week’s press releases. While complexity is a large dragon to slay on the path to enterprise adoption, it won’t stop being a problem after the forward deployed engineers leave the building.
It may reveal itself to be more like a hydra. Every new agent, workflow, and software adds another node in the rapidly sprawling constellation of interconnected services that run a business. Ironically, the very evolutions required for a business to be successful with AI also add kindling to this fire.
As organizations become increasingly dependent on software systems to deliver value to stakeholders, technical debt becomes operational risk. And unlike startups, most mature companies cannot afford to wait six months for newer, smarter AIs to fix the problems they’re creating today.
This is the kind of climate that consultancies will thrive in for years.
That’s all for today. Thanks for reading!
江湖再见🫡
-Sawyer