Please no more personal assistants
They're not the holy grail of personal computing - it's just rich people problems
“Kevin, how come your friend dresses like he’s poor if he’s so rich?” my Sudanese-American mother-in-law quipped.
One of my close friends, child entrepreneur turned Silicon Valley tech magnet, was visiting for my wedding. I personally thought he was looking fly, but rather than attempt to explain the intricacies of Silicon Valley fashion I joked “His clothes are actually very expensive, tech founders just pay to look poor”.
By now I’ve been working on building emotionally rich, soulful, and soul filled AIs for awhile now and I can’t tell you the number of times people have either asked or suggested I just build a personal assistant instead.
I die a little inside each time.
So, I’ve spent countless brain cycles trying to understand: why does everyone in silicon valley seem to think there’s enormous untapped “consumer” market for “assistants”? If you had asked me half a year ago I simply would have been perplexed, but since I understood that social influence was social capital, which can be exchanged for actual capital1, I can now finally explain it to you.
This weird thing happens when you accrue enough influence on the world that inbound connections become more prevalent than need to generate outbound ones - the amplified network of people who you’ve spent enormous effort on cultivating start replicating whatever message you’ve created2. Your email inbox piles up, your twitter inbox, your text messages - one of my close friends gets at least one new SMS literally five minutes and during one of our meetings I couldn’t even think watching her phone buzzing nonstop.
If there’s a surplus of people seeking you, there’s no need to seek, and you have a new problem: choice.
Now, you are acutely aware of how finite your time is, and it begins to fill up with other people’s priorities. You calendar is booked with meetings that other people wanted, your emails have 50 asks a day, you feel the immense pressure of everyone wanting something from you. And so, your new job is to prioritize, sort, and weigh the quality and expected value from each potential interaction, against whatever initial conditions you have set in motion.
It shouldn’t be surprising that at this point, the cognitive pressure of inbound asks begins to dominate your perception of reality, with a totally new expectation that people find you. Your one problem, and in fact, your only problem, is prioritization of inbound information streams.
In order to be effective at inbound prioritization you’ve had to start perceiving people, other humans, as utilities, as tools, as leverage that you rank order. Harboring this belief is almost identically synonymous with financial success, so you are now an executive or a venture partner, you are a business person.
And finally, you’re now in charge of capital allocation.
So let me ask you: what problem do you want AI to solve?
I’m guessing you’ve figured out the punchline by now: Your problems or really, your one problem. Your emails need sorting, your messages responding to, your decisions sorted. You need a personal Chief of Staff (200-500k a year), and if you can’t afford that because they are unbelievably expensive, you need a personal assistant who’s widly expensive (100k a year). If you can’t afford a full time assistant, you can outsource your assistanceship to an assistant outsourcing company for just 30k a year.
Wouldn’t it be amazing if an AI could do all this for you but for 50$ a month! Plus, literally everyone you know happens to have this same issue… You know what?
“Personal Assistants are the Holy Grail of personal computing” you claim loudly in your latest Ted talk.
Whenever you hear a Silicon Valley executive arguing for an AI personal assistant, they are forgetting that their entire way of life, their way of being, is incredibly rare within the population.
I can’t believe I need to say this, but most humans interact with other humans to build, create, and maintain relationships - not to exercise leverage.
In extrapolating their problems, all of their friends’ problems to the general populace, Silicon Valley execs are committing a classic case of false consensus bias.
Is there a viable startup company in here? Sure. Plan to democratize the Chief of Staff (CoS) role. Begin by hiring the best possible CoS in Silicon Valley, make them the face of your AI, and the primary interface to your high net work clients. Target Silicon Valley execs who have enough money to pay 1k a month for a subscription to your service, and only grow your user base as quickly as you can give your CoS leverage through replicating their entire persona and highest quality of service with AI - seriously, this is literally the most obvious financially viable startup opportunity in AI3. I walked into an event and had 200k MRR trying to pay me for this hypothetical product before I left the room.
But honestly, I don’t care about money that much, and what a wildly boring problem to solve anyways - a vanity product to help rich people get richer.
So, no, I am not building a personal assistant.
I recently started a Twitter account after learning about this whole influence thing
I’m not quite there yet but I can see where this is going
Incumbents won’t build it because it only serves rich people. It’s hard to build because requires a network of rich people, and has the highest possible margin of any AI product that doesn’t require regulation. But it’s technically feasible!