The Great AI Hiring Freeze

What happens when writing a job rec is more work than automating a job?

Before I get to my essay today, let me share a preamble on a podcasting author's wonderful, horrible life (page down to skip and read "The Great AI Hiring Freeze").

I decided to start writing again this summer, as I haven't flexed the muscle in years and enjoy the replies I get back.

Also, I'd like to finish the half-done book in my drafts folder. These blog posts/emails are a fantastic warm-up exercise for me, although since I wrote my first book, ChatGPT, and Grammarly have been introduced, and I'm dismayed and inspired.

They fix your typos in real time, restructure your clunky sentences, tell you when a word is overused, and suggest topics. The purist in me feels like some sex coach is over my shoulder telling me, "Now do this and watch what happens!" 

Maybe these things will make you a better lover, but it sure does feel less, umm, organic, if I'm being honest. 

[ Note: It told me to take out the "umm" above for clarity. Fuck you, Grammarly, and I'm leaving it in! Ahh, Jesus. Now it's telling me to replace "take out" with "remove" in the previous sentence, which I am also "dismissing." ] 

As a total aside, podcasting can kill a writer's productivity, huh? 

I peaked at seven podcast episodes a week this past year, with six This Week in Startups and one All In. It's a little work for me, and the payoff is massive, both in advertising dollars and reach (50m+ viewers between both pods every year). 

95% of the founders I invest in tell me they listen to my podcasts, which start the relationship in hour 10 on average -- they know everything about my life, investing philosophy, and world views. They often repeat the advertiser slogans, reminding me of my funniest ad reads. 

When meeting with potential investors (LPs) for my venture funds, the meeting typically begins with discussing a podcast episode and ends with a request for a selfie (which I love doing).

With the massive success of their podcasts, I see much less writing coming from friends Sam Harris and Tim Ferris. We should devise a system for world-class authors where we ban them from podcasting after two years without publishing a book. 

But podcasts are just so perfect, sitting between the crack-hit of a tweet and the ultra-marathon of writing a book. They're mini-audio books or speaking gigs that pay authors $10,000 to $100,000+ (in advertising or sponsorships). 

Compare that to the 12 months of pain to write a book. Or the monotonous, 72-hour travel turnaround associated with corporate speaking gigs. 

Why get on a plane when you can turn on a microphone? 

That's enough preamble; let's get to today's essay. 

The Great AI Hiring Freeze

It's year zero for AI in the workplace, and already it's more work to write some job recs than it is to give that job's tasks to AI.

Need a new conference producer? Write a list of what they do daily and the hours each task takes. Throw those tasks in Google Bard, ChatGPT 4, and Quora's Bard -- all of which are under a year old, COMBINED -- and watch a two-week job finish in two hours of "prompt engineering." 

Do you need someone to write show notes for your podcast or transcribe and time stamp it? Run the Zoom transcript through Notion's new AI processor and watch two days of work finished in two hours. 

Were you considering hiring a Chief of Staff to reply to the bottom half of your emails? I just turned on Superhuman's AI, which analyzes your tone and powers through your emails five times faster -- it's less than a week old. 

Developers and designers are telling similar stories of productivity gains, and while this doesn't feel like you should fire folks and replace them with AI in 2023 and 2024, you should train your current staff to use these tools and go on a hiring freeze.

I'm seeing this in many startups already, and revenue is going up while headcount is staying steady -- which all drops to the bottom line. 

We've Seen This Movie Before

Thirty years ago, when I became an entrepreneur, I watched this trend happen in the IT business, when setting up computers required a "PC Support Specialist" for every 25 to 50 employees. 

I was one, and when you got your PC in the 1990s, I came to your desk and unboxed your computer, plugged in a dozen cables, added some memory chips, and a larger hard drive, installed some software from CD-ROMs, and flipped some jumpers on your ethernet card to "get you on the net." 

Today all of that is abstracted, and if you're under 40 years old, you're laughing at the concept of a "PC Support Specialist." 

Then when the internet came out, startups had to spend their first $250,000 in funding on a co-location facility, $50,000 installing a "T1," and another $250,000 putting down a letter of credit with a landlord for a 10,000 square foot office space. 

All that is abstracted now, with Microsoft Azure giving founders buckets for free cloud credits and remote work. 

You used to have an accounting and HR department; now, that's all outsourced. 

It took 25-50 folks and a year to start a startup in the 90s. Today two to three builders start with < $25,000 in expenses in the first year. 

Age of Austerity

Paradoxically, year zero of the AI revolution arrived at the same time as the great tech and media layoff, and the post-covid standoff between remote workers and management peaked. 

Amazon workers almost returned to work last month, stopping at the doors to, I kid you not, protest coming back to work. The 101 and 280 are still wide open during rush hour here in the Bay, and the luxury buses run no more. 

Tech management can't get folks back to the office, even three days a week, so they're getting their retribution by doing the next best things in their minds: firing managers and low performers, disbanding culture and DEI groups, and instituting hiring freezes.

The hiring freezes, and RIFs demand more from the people who work on the margins, but they also remove annoying distractions from high-performers (which they like).   

These managers are on total tilt reading the stories at the over-employed subreddit and activity stats in Slack/Microsoft Teams.

Some not insignificant number of the employees they hired and coddled over the past two decades are--GASP!--entitled little monsters who weren't working at the office and are indeed working less at home.  

And the results are the results: Twitter didn't go down after the reorg, and Zuck watched his stock price triple when he fired 20,000+ middle managers who refused to return to work. 

This hiring freeze is permanent. 

Management is pissed that they won't come back to work, and since AI is transforming the high-performers into Gods, the solution is obvious: stop fighting and focus on rewarding and being in awe of your 10x employees.

The Benefits, Friends

Of course, if you can't get a job at Microsoft or the Facebook, you're left with three viable options. First, you can become a highly skilled laborer, like an electrician or plumber. 

Second, you can enjoy 3-5x the federal minimum wage that the gig economy companies like DoorDash and Uber are now paying. 

Third, you can embrace AI, become a 10x God, and start your own company, accruing all of those gains to yourself (and maybe your friendly neighborhood angel).

-- end

best, Jason

[ jason@calacanis.com for life; replies come to me ]

PS - I'm not sure if I'll write another piece this week, but I might, depending on how many of my old friends hit reply and share what's going on in their lives and what they think about this piece. This is the best part of writing: the responses from your friends.   

PPS - I was going to put some plugs here but decided against it since y'all know what I'm working on (investing in startups, doing podcasts).