Every week, I write a deep dive into some aspect of AI, startups, and teams. Enter your email address to subscribe below!
Share
New from nerd processor: Which jobs are getting replaced by AI?
Published 9 months ago • 3 min read
AI, you're hired
I'm not a VC, but since the start of 2024, I've looked at pitches or demos from more than 100 early-stage B2B AI startups. That low-level murmur you keep hearing about AI replacing jobs? It's getting louder.
It's not job replacement, silly! It's an AI-driven digital transformation of roles!
Among all the AI startups I've looked at this year, there is a single, abiding theme that they're turning up to 11: They promise they will save you money by replacing your team.
It's not always transparent. The phrase job replacement shows up in media thinkpieces, but it doesn't show up in many pitch decks. Most often, AI startups are pitching job augmentation: ways to extend your current team or make their work more efficient. In many cases, startups pitch their solutions as a cost-effective way to add productivity in an area the customer could not otherwise afford to staff.
Domo arigato, Mr. Roboto!
However, companies are increasingly speaking about job replacement more directly. Across the pitches I've looked at, 38% explicitly compare the cost of their solution favorably to the cost of paying human salaries to complete the same tasks. 76% of them explicitly pitch that, when companies adopt their solution, they will need to hire fewer people.
This isn't brand-new. Some early-stage startups were pitching these same promises in 2022 and 2023. But so far in 2024, I'm seeing it show up in some fashion in almost every single pitch.
A marketer, a recruiter, and an AI walk into a bar
Hey ChatGPT, write me some punchlines for a joke that starts out, "A marketer, a recruiter, and an AI walk into a bar!"
One job that AI isn't replacing any time soon? Comedian.
Among all roles pitched for replacement by AI startups, marketer and recruiter come up the most often, followed by entry-level seller (e.g. Sales Development Representative/SDR) and executive assistant.
Nearly all of the pitches I've seen focus on knowledge worker replacement or augmentation, though those surely aren't the only workplace use cases for AI (and to be fair, this skew is at least partly a byproduct of the B2B pitches that are most likely to come across my desk). There's also a long tail of jobs in the Other category, including paralegals, financial modelers, designers, and more.
What's with the density of marketing, recruiting, and sales jobs being pitched for augmento-replacement in the startup AI pitches? Maybe it's because these jobs all focus on optimizing some kind of content to get a response. Given the right feedback loop, this is something AI does pretty well.
You may be thinking: "But that's not all marketers do! Or recruiters! Or sellers!" And of course you're right. It isn't. In fact, not one of the AI startup teams I've met with personally this year (not 100, but a lot) believes that AI will totally replace these functions even at their own companies.
But as one founder told me, "If you can use AI to automate 25% of someone's job function, you can hire a team that is 25% smaller. That's just math."
A few weeks ago I was talking to a seed-stage investor with a small fund who said they look at 1,800 companies a year to ultimately make ten investments. They aren't an outlier. If those numbers sound brutal, they are.
As much as the math makes it clear, the lived experience of trying to fundraise makes it even clearer and more visceral. Most startups sound a lot like other startups. If you're trying to raise money for a venture that is just getting off the ground, you need more than the perfect team with the perfect technology. You need a market ambition that is huge enough to matter, and you need to turn the expression of that huge ambition up to 11 when you pitch it.
In other words, even startups whose AI solutions just automate a few basic tasks (which is most of them), have every incentive to tell their story in the biggest way possible. And right now, a whole lot of them are telling the story of job replacement.
Party of five Last year, I looked at the growth signals from 180 different startups, and I found that companies with teams of about 5 employees per manager perform best. If you're going to err on one side or the other, it is better to have bigger teams than smaller ones. Today, most companies have too many managers, especially in technology companies. But how will this change as agentic AI enters the workforce? How many managers are enough, and how many are too many? What is agentic AI?...
Ghosty McGhosterson Remember when I collected data from 288 job seekers to figure out who gets called back for the jobs they apply to? On average, people get more recruiter calls when they're in their 30s, working in tech, and looking at in-person roles. Over the last three months, I've gone deep with 30 job seekers who meet this description. These are the people getting the most calls when they apply to open roles. As a group, they have applied for nearly 1,000 roles since October 1, 2024....
nerd processor redux I wrote the first nerd processor on March 5, 2024. Since then, I've published weekly data stories, reflections on my wins and misses as a CEO, and insights about AI, remote work, team communication, management, and lots more. In the last 14 weeks alone, I've published an original data story every single week, complete with custom analysis and graphics, no matter what else has been going on. I've also launched two e-mail courses sharing my playbooks for telling viral data...