On the Socially Acceptable Use of AI in Business

Posted on March 29, 2026 | Leave a comment

There’s a question I’ve been mulling for a while now, and I think it’s time to write it down: when is it okay to use generative AI in a given business context, and when does it cross a line? I’ll focus on two specific areas I know well — board work and strategic analysis — but I think the principles generalize.

Let me start with what I think is the easy part. Using AI to draft a meeting agenda? Fine. Using it to generate a board deck? Also fine, though you’ll probably go to manual edits after the first or second draft. Using it to produce a document summary? Fine [1]. These are tasks where AI is essentially doing the grunt work of organizing information you already possess, and where the human judgment — yours — is the thing that actually matters.

Using AI to produce final documents? That’s dicier today — ask anyone in legal — but I think there’s a simple rule that applies to all of these examples.

That rule? Use AI to do whatever you want, but you own the output. Not the AI. You. If it’s wrong, that’s on you. If it misses something important, that’s on you. The moment you present something to a meeting, a customer, or a board, you are vouching for it. Saying “well, AI generated that part” is not a defense. It’s an abdication of duty.

The “Ad Hominem” Problem

Here’s something that bothers me about the discourse around AI-generated content: people hear, or even suspect [2], that AI wrote something and it’s immediately dismissed — not because of anything wrong with the content, but because of how it was produced. That’s a logical fallacy. Specifically, it’s a variant of ad hominem: attacking the source rather than the argument [3].

I frequently need to remind people of this. Judge what was said, not who — or what — said it [4] [5]. If the analysis is sound, the framing is useful, and the questions raised are the right ones, then the mechanism of production is largely beside the point. The quality of the thinking is what matters and what should be challenged.

That said — and this is important — the inverse is also true. Producing AI-generated content and presenting it as your own thinking is not okay. The problem isn’t that AI helped. The problem is the pretense that you did the thinking when you didn’t. Ownership means you’ve read it, challenged it, corrected it where it was wrong, and can defend it. If you can’t do that, you haven’t done your job.

AI as Calculator

I’ve always thought the right analogy for AI is the calculator. A wildly more powerful calculator, obviously, but a calculator nonetheless.

When calculators became ubiquitous, people lamented the loss of slide rule proficiency. And yes, something was lost. But the point of mathematics was never arithmetic. It was reasoning. If the calculator handled the arithmetic error-free, you could spend more time on the part that actually matters. The same logic applies here: there’s a lot more to argument and strategy than copywriting or slide formatting. If AI can handle the scaffolding, you should be able to spend more time on the substance.

The complication — and it’s a real one — is that AI can start to approximate thinking in ways a calculator never could. A calculator doesn’t write your memo. It doesn’t suggest your strategy. It doesn’t synthesize twenty pages of board material into five crisp questions. AI does all of that. And that creates a temptation toward laziness that calculators simply didn’t. The laziness is the problem, not the temptation toward it.

There’s also research starting to emerge suggesting that relying too heavily on AI can actually impair your own reasoning. You offload the synthesis, and you stop synthesizing. You offload the framing, and you stop framing. The cognitive muscle atrophies.

I was not surprised when I read reports that people with long streaks in Duolingo couldn’t speak well in practice. In my view, as a half-decent French speaker: if it doesn’t feel like work, you’re probably not learning [6]. Corrolary: if it doesn’t feel like work, you’re definitely not working.

How I Use AI in Board Work

Here’s what I often do with AI today. I sit on several boards, and I’ll sometimes load a board deck into a generative AI tool before the meeting. I ask for a summary. Then I’ll ask how it thinks the company is doing. I’ll then ask for the top 5 questions to ask in the meeting. Then, I’ll go read the deck with an eye toward what’s been extracted [7].

And then I’ll go back and challenge the AI. I think issue three is more important than issue one. I think it missed issue seven totally. I think issue two isn’t an issue; the company’s fixed it already. Often, I’ll bring competition into the picture because (in my humble opinion) most boards don’t spend enough time thinking about competition [8].

And here’s the question I’ve been wrestling with: should I be transparent about using AI to help generate those issues (or questions) when I bring it to a board meeting?

My instinct is yes. If I want to send the CEO a list of top five issues facing the company before the meeting, I have two choices:

  • I can pretend I wrote it myself, unassisted. Complete with typos and hyphens instead of em-dashes.
  • I can say, “here’s what I generated with Claude after iterating on your board deck” and copy/paste the final transcript.

Now, I know what management can think: “Well, we could have asked Claude, too” [9]. And I’m okay with that. My response would be: “Well, then, why didn’t you?” I just want the best topics list.

To me, the question isn’t where the list came from. The question is whether it’s the right list. That’s the only question that matters. Boards have very limited time together. We should think hard and use all available tools to ensure that we’re spending that time on the right issues.

The point isn’t the slides or the questions list or the agenda or the summary. The point is the conversation. To maximize value, we need to be having the right conversation. No talking about things easy to talk about. Not going through the motions. Not death marches through templates, much as I love both templates and death marches.

This takes me back to calculators. I can check the math on your board slides using pencil and paper. Or I can use a calculator. Or I can upload your table and ask Claude to check the math. We can take a test with our calculators secretly on our laps or with them in plain sight on our desks.

I vote for the second option. Use all available tools. Don’t use them clandestinely. Use them out in the open. But don’t abdicate to them. Own the output. This isn’t Claude’s list of our top five challenges. It’s my list, built using Claude [10]. Better yet, it’s my list, period. (But I’m not going to hide that I used Claude to help build it much as I wouldn’t hide that I used a computer and a keyboard.)

That’s where I am. I’m curious where you are. Is there a line you’ve drawn in your own work? Do you think transparency about AI assistance is a norm we should be enforcing, or are we creating a two-tiered standard we’d never apply to other tools? Let me know in the comments.

# # #

Notes

[1] Just as long as you also read it or are prepared to say, “I didn’t read it, I only read a summary.” FWIW, I find it useful to generate a summary, read it, and then read the document. And sometimes, then go back to the summary. The summary ends up serving as a reading guide.

[2] Thanks to tells like the dreaded em-dash.

[3] I had to air quote ad hominem because — thanks to my high school Latin teacher, Mr. Maddaloni — I know that ad hominem means literally “toward the man.” There is thus not only gendered language (heck, it was nearly 3,000 years ago) but considerable irony in speaking of ad hominem attacks on a machine. Ad machina, anyone?

[4] By the way, this is the exact opposite of most social media behavior.

[5] This isn’t just good intellectual hygiene — it’s a reliable way to reduce or eliminate bias. When you evaluate an argument on its merits rather than its source, you sidestep a whole class of distortions: the tendency to over-credit ideas from high-status people, to dismiss ideas from unexpected sources, or to reject a perfectly sound analysis because you don’t like the messenger. It’s a discipline worth practicing whether the source is a junior analyst, a competitor, or a language model. The argument either holds up or it doesn’t. That’s the only test that matters.

[6] This is a critique on gamification, but also is highly related to the topic of customer value metrics, about which I’ve written with my Balderton EIR colleague Dan Teodosiu.

[7] By the way, the wordier the board deck, the more this process helps.

[8] This itself could be a long discussion but remember three things: my first job in marketing was competitive analyst; I believe strategy is either “the plan to win” (Burgelman) or “the way to overcome our biggest challenge” (Rumelt), and ergo it cannot be done without looking at the market. North Stars are great, but they don’t tell you about the army you’re going to face when you hit latitude 55 degrees 45 minutes.

[9] And this is probably the kindest thought. Others might include:

  • Perfect — now the monkeys have flamethrowers
  • Fantastic — it’s like giving toddlers espresso and a whiteboard
  • Great — now the VCs can skip even faster to the wrong conclusion
  • Terrific — now it’s gut feel with citations
  • Right — so now we’re pattern matching with turbo-autocomplete

(And those are manual em-dashes.)

[10] I assume that we are not all going to have the same AI conversation or all use the same tools. The way I push Claude is going to be different from the way another person does.

Leave a comment

Posted in AI, Boards, Management, Silicon Valley, Startups, VC

Tagged , ,

The Odd Little Book All Founders Should Read On Selling Their Company

Posted on March 27, 2026 | 2 comments

I recently read The Magic Box Paradigm by Ezra Roizen. It’s self-published, was first released in 2016 [1] , and you won’t find it on most startup reading lists. The writing is uneven and inconsistent. The metaphors are weird. There are too many TLAs (three-letter acronyms). Nevertheless, I think all founders should read it — early and often. Early, meaning years before you contemplate selling your company; often, because if you read it early, you’ll need a periodic refresh.

Everyone in M&A has heard the expression “great companies are bought, not sold.” It gets knowing nods in board meetings, but is then promptly ignored in practice. The reason it’s hard to internalize isn’t that the idea is obscure — it’s that acting on it requires you to behave in ways that feel completely wrong. It requires you to slow down, stay deliberately vague, and resist the urge to pitch. For a founder who has spent years getting good at pitching, that turns out to be genuinely difficult to do. Knowing something and behaving consistently with it are two different things. [2]

The Magic Box

Roizen’s central metaphor is the “magic box.” Some things are popsicles — they have known, comparable value, and you can auction them with reasonable confidence in the outcome. Startups are not popsicles. They’re magic boxes. A startup’s value isn’t fixed or objectively discoverable; it depends almost entirely on who’s opening the box and what they plan to build once they have it. The same company can be worth $50M to one acquirer and $500M to another — not because of negotiating leverage, but because of genuine strategic fit. Which means the job isn’t to run a wide process and let the market discover your price. It’s to find the buyer for whom your value is highest and help them see it — ideally before you ever hire a banker.

The Trail of Tears

The best part of the book is the start of Chapter 6, describing what Roizen calls the sad path. I call it the Trail of Tears, because founders walk it constantly — and the thing that makes it a tragicomedy is that every single step feels reasonable at the time.

These two pages are worth the price of the book alone (highlighting mine).

Tragicomedy.

For those who can’t read the images, it goes like this. One of startup Alpha’s investors happens to meet GiantCo’s head of corporate development at a conference. Corpdev thinks Alpha might be worth a look. The investor, delighted to add value, makes an introduction. Value added! [3]

The next day Alpha’s CEO gets an email from Corpdev asking for a deck he can socialize with the relevant product teams. The CEO panics slightly — what do I send? — and settles on his most recent investor presentation. It worked great for raising a big round. It’s got product detail, market sizing, competitive positioning, go-to-market strategy. Should do just fine.

Corpdev reviews the deck. He sees some potential but no clarity on where Alpha’s products might fit into GiantCo’s portfolio. He forwards it to a few product leads and a general manager. The deck is salesy. It was designed to solicit investment in Alpha as a standalone company. The deck’s salesy quality is read by GiantCo as a sign that Alpha is trying to sell itself.

A presentation is scheduled. The relevant GM — probably the best potential internal champion — can’t make it. The demo goes well. GiantCo’s attendees are engaged. The meeting ends with enthusiasm and a commitment to follow up.

Corpdev follows up by sending Alpha’s CEO a list of boilerplate diligence questions: financials, cap table, customer concentration, licensing. The kind of get-to-know-ya questions that corporate development types like. Roizen’s line here is worth remembering: Corpdev is using an X-ray when a telescope is what’s needed. Alpha’s CEO, under pressure from his investor for updates, has his finance team pull together a packet in response. Everything is proceeding mechnically at this point.

Corpdev now takes a critical look. Revenue is concentrated. Burn is high. Valuation expectations are probably rich given the cap table. Alpha is too early and GiantCo is too busy. The eventual reply: we really like what you’re doing, but it doesn’t map to any current priorities. Let’s stay in touch and try to connect again at next year’s conference. In short, you’re a nice guy/gal, but let’s be friends.

The investor wants an update. The CEO has to explain that nothing happened.

The wrong deck. The wrong follow-up. The right GM missing from the meeting. An X-ray instead of a telescope. No chance for the idea to form inside GiantCo. And now the bad news needs to be broken to an investor who was just trying to help.

It’s a sad path indeed. And the reason it works so well as a teaching device is that the CEO didn’t do anything stupid. They made reasonable calls at every step. That’s the point.

This literally happens every day. It wastes time. It’s demotivates founders, raising and then dashing expectations. Worse yet, its leaves the company with a residual “those guys are for sale” taint — a mark that’s hard to see and even harder to erase. [4]

The Partner Big Idea

What should have happened instead? Roizen’s answer is what he calls the Partner Big Idea (PBI). The mechanics of building a PBI are more involved than I’ll go into here — read the book — but the core principle is this: the deal has to become their idea, not yours.

The investor presentation was the original sin. It accidentally signaled that Alpha was for sale, which put GiantCo in evaluation mode rather than strategy-building mode. What Alpha needed wasn’t a buyer to evaluate it. It needed a champion within GiantCo — ideally that GM who missed the meeting — to develop a strategic vision that Alpha was necessary to execute. Not “Alpha is an interesting acquisition target” but “here’s a thing that we need that we can’t build without Alpha.”

Building that requires a totally different set of behaviors. It means getting to the right person quickly — the GM or product leader whose roadmap would actually change — and not spending lots of time with Corpdev. It means asking more questions than you answer. It means leaving the story incomplete enough that the other side has room to build it with you. Incompleteness, in this context, is a feature. It gives the champion something to build and own. [5]

Corpdev is not that person. Corpdev manages process and filters opportunities. They can help once a deal is real, but they rarely create the reason for the deal to happen. If your primary relationships are with Corpdev, you’re operating inside a system designed to evaluate, not to originate. And it’s a system that, left to its own devices, will evaluate your company on a financial, not a strategic, basis.

An Investment Banker Weighs In

I asked an investment banker friend, who works regularly with top strategic buyers, about the book and its relevance today. He had three key observations.

First, the importance of partnerships as a precursor to M&A has only grown since the book was written. Companies partner, integrate products, share customers. Over time, the relationship gets embedded in each side’s roadmap. At that point the “big idea” isn’t hypothetical — the buyer doesn’t just believe in the opportunity, they depend on it. The magic box becomes a dependency.

Second, geography matters to some more than the book acknowledges. Snowflake, for example, drew a reasonably hard line for a long time: a deal couldn’t happen unless the technical team relocated to one of their engineering hubs. With return-to-office (RTO) continuing to gain momementum, I think this will continue to increase in importance.

Third, don’t underestimate the importance of team buy-in. Strategic acquirers aren’t just buying code, they’re buying the team that builds it and they can tell the difference between teams are cashing out (and who will work until exactly the day their handcuffs disappear) and teams who are genuinely excited about a combined future. As a reflection of this, buyers are increasingly splitting the payment, sending more money to the retention pool and less money to the cap table. This creates a tension between investors and employees, but it all gets negotiated in the process.

But What About Banker-led Processes?

At this point, you might reasonably ask: how does all this square with the standard advice to hire a banker and run a process? Aren’t these two ideas in tension?

Not really. They operate on different timelines.

The work Roizen is describing is long-term and strategic. It’s about shaping how a potential acquirer sees your company years before any process begins — helping the right person inside the right company build a strategy that depends on you. You’re not selling the company. You’re teaching someone else why they might need to buy it. The Magic Box Paradigm is about getting bought.

A banker-led process is something else entirely. It’s about getting sold. It’s a short-term mechanism to create urgency, surface alternatives, and establish price. It can accelerate a deal. It can’t create the underlying reason for one to exist.

If the strategic groundwork has been laid — if there are multiple potential acquirers who already “get it” — then a process can work extremely well. It forces those buyers to act, on a timeline, in competition with both PE sponsors and one another.

If the ground hasn’t been laid, then the process tends to default to financial evaluation. You get Corpdev questions, lukewarm interest, and a lot of “not a priority right now.” In other words, a scaled-up and more formal version of the sad path.

One nuance here, having lived it: the hardest part is aligning timelines.

A PE-led auction runs on a clock. You set dates, people prepare bids, and the banker’s job is to keep everyone moving in a tight, predictable cadence. That’s how you create urgency and price tension.

Strategics don’t work that way. They need time — to line up a champion, to socialize the idea internally, to get product, finance, and executive buy-in. Occasionally they can turn on a dime, but that’s the exception, not the rule.

The tension is obvious. Run the process too fast and you lose the strategics. Run it too slow and you lose the auction dynamics. This is why you need to have relationships in place with strategics well before your banker process begins. Otherwise they simply cannot keep up.

The banker’s real job, in this context, is to try to align those timelines. Because the worst outcome is hearing what I once heard: “We’re very interested, but we can’t possibly execute on that timeline, so we’re going to drop out.”

And once that happens, you’ve lost exactly the buyer who might have valued you the most. Utter process failure. Think: You had one job!

So the two ideas aren’t in conflict. They’re parallel. Do the strategic work early — years before you’d contemplate selling. Then, if and when the time comes, use a banker to run a disciplined process on top of it.

Read it Early

This is not a book about how to run an M&A process. It’s a book about how deals actually form — which is a different and more important topic. The sad path exists because most founders don’t think about this until they’re already in it, at which point it’s very hard to correct.

Read it at least four years before you think you need it. Let it shape how you build relationships with potential acquirers. Help the right person inside the right company build a strategy they can’t execute without you — and make sure they realize it before you ever hire a banker.

If you do that, you may not need a process at all. And if you don’t, you can’t count on a process to save you.

# # #

Notes

[1] With a second edition published in 2023

[2] This is actually a broader problem in business. The list of things people nod at in board meetings and then promptly ignore would fill several books.

[3] Roizen’s deadpan “Value added!” is one of the funnier lines in the book.

[4] To be clear, the taint is that they’re always for sale and nobody wants to buy them. Imagine the house on a street with a perennial for-sale sign in front of it.

[5] This is counterintuitive enough that it’s worth sitting with. The instinct is to show up with a complete, polished narrative — that’s what pitching trains you to do. But a complete narrative leaves nothing for the other side to build. Their investment in the idea comes from the act of building it.

2 Comments

Posted in Entrepreneurship, Leadership, M&A, Management, Silicon Valley, Startups

Tagged , ,

Be the CMO Everyone Wants to Work With

Posted on March 21, 2026 | Leave a comment

I’m a big fan of Dave Gerhardt and the Exit Five marketing community he has built. While I have always liked the idea of peer-networking communities, I think they change from vitamin to painkiller in times of rapid change. Why? Because the playbooks haven’t been written yet.

Take SaaS, for example. Today, you can find scores of blogs — including mine — that talk about how to run SaaS businesses, how to plan SaaS businesses, and how to produce and interpret SaaS metrics. Twenty years ago virtually none of that existed. You needed to figure it out. And one of the best ways to figure it out was to spend time with other people who were doing the same figuring.

There’s a time for timeless wisdom and there’s a time for talking to other people who are doing the same thing as you are, right now. I think AI and the massive disruption it creates in marketing — from content to workflow to performance marketing to analytics — means it’s an awesome time to join a marketing peer-networking community.

I tell this to literally every CMO I work with:

  • Come to me for timeless wisdom (if not fleeting opinions).
  • Join Exit Five (or equivalent) to talk to peers who face the same challenges you do, every day.

That’s not to say that following thought leaders isn’t a great tactic, too. I keep an eye on Emily Kramer, Elena Verna, Carilu Dietrich, Alice de Courcy, and Jon Miller. And I can’t wait for Rand Fishkin’s rumored new book on Zero-Click Marketing. If there are other marketing thought leaders you think I should be following, please let me know.

All this is why I was thrilled when Dave Gerhardt invited me to speak at Exit Five’s Marketing Leadership Retreat on March 19-20 in Phoenix. Given the above, I knew I wasn’t going to be the person delivering fresh-from-the-trenches information on AI tools and methods. The audience is 100x more qualified than I am to do that.

So what did I want to offer up instead? Some timeless wisdom. Specifically, timeless wisdom not just on how to successfully do the CMO job, but on how to be the CMO everyone wants to work with.

Why that topic? And bear in mind it takes a lot for me to pick a title that ends with a preposition. Because I thought it captured the key to success in an important way.

  • It’s not just about doing the job. Yes, that’s quite hard already but if you only focus on that you ironically increase your odds of becoming a statistic.
  • It’s not just about keeping the job. And yes, there’s an art to that, a big part of that is simply remembering to market marketing.
  • It’s about helping the boss do their job which not only increases your strategic value, but makes you more “sticky” in your role.
  • It’s about doing all that while being the CMO that everyone wants to work with (EW2WW)

Why does that last point matter?

  • Marketing is inherently a service organization, so a strong internal customer service orientation never hurts.
  • Being the CMO EW2WW helps you keep your job. With a median tenure of 18-24 months, this should never be too far from a CMO’s mind. Belt and suspenders.
  • If everyone wants to work with you, it helps you find your next job. Board members and recruiters — your top two job sources after peers — will seek you out when the time comes.
  • It might well help you get promoted to COO or CEO. If everyone wants to work with you, they might well give you a shot at the next level.
  • It frames things as a positioning problem. And we know a lot about solving positioning problems, working backwards from a desired result.

Ultimately, I’m saying that CMOs should want to position themselves as the CMO EW2WW.

Put differently: “Marketer, position thyself.”

(Adapted without permission from Luke 4:23.)

With that as background, here are the slides from the presentation, embedded below, and downloadable here. I had a fairly miserable time in Gamma building them so apologies for the upside-down funnel, some of the formatting, graphics, and mechanics (e.g., the absence of copyright notice and slide numbers). There is only so much time I’m willing to spend explaining to a chatbot what expression to put on the reflected image of a face in a mirror. And once you drop into PowerPoint to make changes, there seems to be no going back.

Thanks again to Dave for having me and to Allison Saxon for working with me to make it happen.

Leave a comment

Posted in AI, Career, Communications, Leadership, Marketing, Silicon Valley, Startups

Tagged ,

Why I’m Not Worried About Running Out of Work in the Age of AI

Posted on March 19, 2026 | 8 comments

When the auto industry was being decimated and jobs offshored to Japan, many of us instinctively reached for our college economics textbook and started thinking about fish and coconuts — the example Paul Samuelson used to explain David Ricardo’s theory of comparative advantage in international trade.

Well, if Japan is better at making cars than we are, then they should make cars and we should do other things. It’ll be rough for the displaced auto workers but it’s merely a transition period and, like Ricardo says, it’ll be a win/win for both countries and we’ll all be better off in the end.

If you put aside the many concerns with basing policy solely on comparative advantage (e.g., national security, supply-chain resilience, support for fledgling industries, externalities), you could at least get a good night’s sleep.

After all, as I pointed out in my 2026 predictions post, while it’s painful in the short term, we typically don’t miss the jobs displaced by technology.

I don’t think anyone misses elevator operators, a job that peaked in the 1950s and is all but extinct today. The same is true of switchboard operators, pin setters, lamplighters, linotype setters, icemen, typists, shorthand secretaries, and slubber doffers. Yet all these jobs were automated away. While I’m sure the transition was rough for any given elevator operator, as a society we don’t miss them. Most coal miners work hard so their children won’t have to be coal miners.

It sucks for them — the displaced workers — of course, but collectively it’s good for us. Or, in econ-speak, trade creates net national gains, but the costs are often geographically concentrated and painful (e.g., the Rust Belt). If you want to know what it feels like to be on the other end of this displacement, listen to some Bruce Springsteen: Youngstown, My Hometown, Independence Day, and my favorite, The River. That’s what it feels like in a dying town. That’s the darkness on the edge of town.

The pain is real and while occupational transfers are a solution, some are easier than others. An autoworker could transition to another manufacturing job (though often at lower wages) but would be unlikely to become a mechanical engineer. For the economy in general, jobs transition from one sector to another. For any given person, one displacement could be the end of the line.

Another excerpt from the predictions post:

Let’s remember where the word Luddite comes from. Today, we use the word to describe people opposed to new technology. But the Luddites were part of a labor movement; they didn’t destroy machines because they were technophobes, they destroyed them to protect jobs. So the transitions can be rough — and come with social unrest — when jobs are automated away. But we still don’t miss the jobs they fought for.

So, with our safe emotional distance from a 1970s auto worker, what would we have wanted to tell them at the start of the disruption? I’d say:

  • This isn’t a blip, it’s a trend. This is not a problem you wait out.
  • You could try to fight your way through this, but the odds will not be forever in your favor.
  • If you’re nearing retirement, ride this one to the beach and go retire.
  • If you’re early or mid-career, you need to switch horses.
  • Switching early is better than switching late. The goal is to get onto the next horse before everyone else realizes the race has moved.

The popular horse-switching fantasy answer is retraining. “Go back to school and become an engineer.” In theory, yes. In practice, rarely. The jump from an assembly-line worker to an engineer requires years of schooling and a different educational foundation.

The real move for the autoworker was sideways, not upward: industrial maintenance, tool and die work, welding, industrial electrical work, construction trades, trucking, or logistics.

None of these are glamorous answers, but they share an important property: they’re adjacent to the skill set the person already has. They’re realistic.

And now, without the emotional distance we have from a 1970s auto worker, let’s try applying the same thinking to ourselves. One way to do that is to imagine what someone from the year 2070 might write back to us.

They might start the same way.

This isn’t a blip, it’s a trend. Don’t assume this wave of AI capability will conveniently stall just below your job description. You can try to fight your way through it, but the odds will not be forever in your favor. The job you’re defending may still exist, but the number of people needed to do it may fall dramatically.

They might add something else we don’t particularly like hearing: if you’re late in your career, you may be able to ride this out. Large organizations change slowly and adoption curves usually take time. But I think less so in this case — it might prove to be a short ride. If you’re early or mid-career, you should be thinking about switching horses.

But here the analogy gets interesting.

The auto worker in 1975 often had a high-school education and a skill set tightly tied to the factory floor. The realistic moves were sideways into adjacent trades.

The knowledge worker today, by contrast, is already highly trained. Many have college or graduate degrees and have spent years building transferable skills: writing, analysis, design, coding, planning, advising.

In other words, knowledge workers should be much easier to redeploy than auto workers ever were.

So yes, this all hits different because this time it’s us. That’s where we lose the emotional distance, and why the people trying to scare us sound so convincing.

Citrini’s recent thought experiment imagines a negative feedback loop with no natural brake. As they describe it:

AI capabilities improved, companies needed fewer workers, white-collar layoffs increased, displaced workers spent less, margin pressure pushed firms to invest more in AI, AI capabilities improved …

Repeat.

In that scenario, the problem isn’t sudden unemployment so much as a slow erosion of the income base that supports the modern consumer economy. Highly paid white-collar workers lose their jobs, move into lower-paying work, compress wages across the service sector, and reduce spending at the same time.

The one thing I find most missing from the AI future analyses I’ve read is a simple realization: in my experience, there is always, always more work yet to do.

This relates to how I’ve always thought about the total available market (TAM) for business intelligence (BI). Over the decades I’ve worked in and around BI, I’ve been asked countless times whether the market was saturated or saturating.

My answer was always the same.

One day I’ll go to a cocktail party and ask everyone in the room: “Does everybody have all the information they need to make data-informed management decisions?” If everyone says yes, then on that day the BI market will be saturated.

But that day strikes me as a long way off. It did 40 years ago and it still does today. We have accomplished so, so much in BI and analytics. But there is still so much more to do.

Since no AI post would be complete without a reference to Jevons Paradox, let’s invoke it here: When technology makes something cheaper or more efficient, we tend to use more of it, not less. When steam engines became more efficient, coal consumption didn’t fall — it exploded because steam power became economical in far more applications. The same dynamic has repeated itself with computing, storage, and bandwidth. Every time technology dramatically increases our ability to do something, we don’t run out of work. We discover many more things worth doing.

The vast majority of management and strategic consulting today is about focus. And the reason focus matters so much is simple: execution bandwidth is limited. Organizations can only get so many things right.

So managers spend an enormous amount of time cutting, cutting, and cutting. Prioritizing, prioritizing, and prioritizing. Literally speaking, if we had to list the one key to business success, it’s focus. Why? Because there is so much work that we could do. So many strategies that we can imagine. So many things that we could do to help our customers.

It’s not that there’s only a little more work to do. There is an unfathomable amount of work yet to be done. The job of management is pruning that list down to something the organization can actually execute.

Which is why the idea that we’re somehow going to run out of work strikes me as absurd. It feels like a theory written by people who haven’t actually spent much time doing the work in the first place — serving customers, building products, and running businesses. There is always more that could be done.

There’s also an irony here. Increasingly, the people who say SaaS will stand for “service as a software” are the same people saying AI will massively disrupt employment. But when you sell service-as-a-software you are definitionally in the solutions business, solving problems, not merely delivering data and algorithms. And when you look at your customers through the lens of “problems yet to be solved,” you quickly discover that the list is very, very long.

So no, I’m not particularly worried about us running out of work to do.

I do think the transition could be difficult, especially if you’re not ready for it. So, given everything we’ve said thus far, here’s the advice I’d give to a knowledge worker heading into this transition.

You want to be driving the tools, not driven by them. Aggressively learn AI. Be the person who knows the most about solving problems using AI tools — integrating them, automating workflows. Not just generating content.

You want to understand the business outcome of the work you do. For example, if you’re in public relations, your end goal is awareness. Learn how to use these tools not just to generate articles but to drive awareness through whatever channels and mechanisms will make that possible.

You want to know a little about how the tools work, but not too much. People like me are naturally drawn to understanding the internals, but given the complexity of modern AI systems the more practical skill will be knowing how to use them.

One profession today that looks closest to the autoworker analogy is copywriting. If you’re a copywriter, you should already be learning multimedia and learning how to use AI tools to generate and manage all forms of content. Five years from now you won’t be manually writing and editing. You might be checking work, guiding it, generating variations, and certainly personalizing it at scale.

I know it’s a cliché, but I’ll end with a famous line from Wayne Gretzky, who attributed his success “not to skating where the puck is, but to where the puck is going.”

The puck is going toward heavy AI automation. That will create a new class of jobs around using AI to automate work — and another layer above that focused on managing, directing, and orchestrating those systems to actually run and grow businesses.

Skate there.

8 Comments

Posted in AI, Disruption, Management, People

Tagged , ,

The Silicon Valley Canon, Circa 1998: A Stroll Down Software Memory Lane

Posted on March 15, 2026 | 6 comments

Something fun happened today. A reader reached out who had been digging through my early-2000s and 2010s posts trying to understand the history of the software industry. That immediately got my attention because I love people who study history. It’s the best way to understand the present. And a great way to avoid repeating the mistakes of those who preceded you.

So I’m always happy when someone wants to talk about software history.

His specific request was interesting: he was looking for case studies or books that were popular at the time — something that would help him understand how people in the industry were thinking back then.

I decided to do him one better. In my view, the real canon of books that shaped enterprise software thinking was largely written before 2000. So I assembled the following reading list: a set of 1990s-era books on software, strategy, marketing, and the industry itself that many of us were reading while the enterprise software industry was taking shape.

Think of it as a reading-list stroll down software, and Silicon Valley, memory lane.

1990s Era Tech Marketing and Strategy Books

High-Tech Marketing

Positioning / Marketing Foundations

Technology Strategy / Innovation

Product Marketing Culture

Enterprise Sales / Go-to-Market

Economics of Software / Networks

Enterprise Technology Industry Case Studies

Software Engineering

Classical Strategy

6 Comments

Posted in Enterprise, Entrepreneurship, Leadership, Management, Marketing, Silicon Valley, Startups, Strategy