The Idiot Index for Code

Why building is no longer the hard part - a guest post by The Leverage’s Evan Armstrong

💡 This Week’s Big Ideas

💡 Troy Kirwin says there are only three things that matter for a pre-PMF B2B founder to get out of their investor (beyond capital)

🤖 Fareed Mosavat says AI has made him fall in love with coding again

🔮 Jon Lai argues that most investors misunderstand the Genie 3 launch

💼 Join our talent network for more opportunities. If we see a fit for you, we'll intro you to relevant founders in the portfolio

Today’s essay is a guest post by Evan Armstrong, founder and CEO of The Leverage, a newsletter about shifts in the tech market.

Evan argues that the capabilities unlocked by tools like Claude Code, Cursor, and Codex call for a radically different approach to prioritization. We hope you enjoy it.

“Cocaine in the form of code” is how one founder recently described the experience of using Cursor to me. Another called Claude Code “his favorite video game.”

You’ve probably felt it right? There is something wildly addicting about coding agents—the sense of power is heady, each new command sends a rush of electricity to the fingertips, a feeling that you can make anything.

This is a trap.

The capability explosion in AI coding agents is real. Claude Code, Cursor, Codex—they’ve collapsed the time from idea to working software from weeks to hours. The skeptics were wrong. You can build internal tools, dashboards, workflow apps in an afternoon. It’s a new age for startups.

But the interesting question isn’t “can you?” It’s “should you?”

Build, Buy, or Delete

The classic question in operations is build versus buy. Do you take on the work yourself or outsource it to the market? Resistance to outsourcing has killed countless companies who insisted on vertical integration. The smart move is usually to own whatever “makes your beer taste better” aka differentiates your product, and buy everything else. You’ll never become a billion dollar company on the basis of your HR software, so just buy it from someone.

For early-stage startups, there’s a third option that’s almost always wiser: do without. Coding agents have made building so frictionless that we jump straight past the important question—does this process need to exist at all? Build-versus-buy assumes the thing should exist. Often it shouldn’t.

In Walter Isaacson’s biography, Elon Musk describes the algorithm he drills into SpaceX and Tesla teams:

1. Question every requirement. Every choice should have a real name attached to it. Then, you should question every single one of these assumptions, no matter who made it.

2. Delete any part of the process. Complexity is the death of speed and is anti-innovation. You should remove things to the point of failure. As Isaacson wrote, “If you do not end up adding back at least 10% of them, then you didn’t delete enough.”

3. Simplify and optimize. Only do this after deleting. Deletion is a precursor to improvement.

4. Accelerate cycle time. Again, speed is what matters. Increased speed greatly improves the flow of capital. But it should only follow step three, as there is no use accelerating processes that will end up being deleted.

5. Automate. Only once everything has been questioned, deleted, simplified, and accelerated can you automate. Musk learned a painful lesson when he tried to automate Tesla production too soon, leading to the “most hellacious period of his life” where he had to rip out expensive robots in favor of human labor.

The trap is that coding agents make Step 5 feel like Step 1. Building is so easy now that founders skip the prior questions. “Surely I can ditch my CRM? It feels like it is overpriced and isn’t customized to my needs. How hard can it be to make? After all, it’s just a database.”

(Famous last words.)

Instead, Elon’s algorithm would be asking: Can I delete this workflow entirely? Do I need a CRM right now or can I just make do with a Notion database? Can I tolerate a manual process? Is there something off-the-shelf that gets me 80% there for 30 bucks?

This is the danger of agentic coding. Making software cheap/easy to make does not mean you should use that capability replacing SaaS subscriptions willy-nilly (at least in the early stages). To understand why and what you should be doing, we return to Elon.

The Idiot Index for SaaS Has Inverted

Musk uses something he calls the “idiot index”—the ratio of a component’s cost to its raw materials. A $1,000 part made from $100 of aluminum has an idiot index of 10. “If the ratio is high,” Musk says, “you’re an idiot.” The delta is a waste of overcomplicated design, bloated process, or someone extracting margin from your ignorance.

For decades, the idiot index on software was brutal. Engineers were expensive. Code was hard to write, harder to maintain, and nearly impossible to deploy. The underlying primatives—logic, conditionals—were free. The finished product cost a fortune.

AI has started to invert this. So when code becomes cheap, what becomes expensive?

For early-stage founders, the answer is obvious. It’s your time.

Attention is the new raw material. Every hour in Claude Code building a custom CRM view is an hour not talking to customers, not closing deals, not iterating on the product that actually matters. The idiot index now measures how much founder time you’re burning on internal tooling that some Google Sheets and duct tape could handle for $7 a month.

And it’s not just the build. It’s everything after. The tool becomes a maintenance orphan, built by a coding agent and owned by no one. Then there are also the context-switching costs. Whoever you hire will be familiar with the basic productivity suite, or with tools like Notion or Airtable, but they will be baffled by whatever vibe coded alternative you came up with.

All of these add up to death by a thousand cuts, micro-distractions from doing the only thing that matters before your Series A: achieving product market fit. Everything else is pointless.

So before opening Claude Code, ask:

1. Does this need to exist? Many workflows don’t for the first year.

2. Can I do it manually at 70%? Good enough is good enough pre-PMF.

3. Is there a $30/month SaaS? The subscription is always cheaper than your time.

4. Who maintains this in six months? If the answer is “the founding team”, that’s a cost.

5. What else could I do with these two hours? If the answer is “talk to customers,” you’re done.

The founders talking to me about their addictions to coding agents feels like the perfect metaphor. Drugs give you nothing but dopamine, while coding agents give you dopamine and a deployment.

The feeling of progress plus the artifact of progress, minus the actual progress. That’s the worst trade a founder could make.

---

Like this post? Forward it to your team! And be sure to subscribe to Evan Armstrong’s newsletter, The Leverage.

For more weekly dives into the world of early stage startups, subscribe to our newsletter below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.