We all need better personalities (asap)
The latest release of OpenAI’s ChatGPT will re-frame what it means to be creative as individuals and cultivate leverage in business models
Read the AI-generated summary of this blog post here. IT’S INSANE.
I may have just “wasted” a rare hot weekend in Melbourne talking to a chatbot.
For those who missed the headlines, OpenAI just released their ChatGPT tool which uses natural language processing to generate everything from stories, movie scripts, marketing campaigns, code, poems, business plans…you name it, literally.
If you’ve been trawling through Twitter recently, you’ll know that nearly every second person was a having a existential crisis about their careers. As did I. For the past 36 hours, I’ve questioned my value as a “creative” — heck, more like my value as a person with a distinguishable personality.
There are some great examples of people using ChatGPT here, but I had way more fun trying my own prompts…
Like, do I even have a job now?
Sure, the concept of AI has been around for a while, but with ChatGPT being so openly accessible and creepily creative, the impacts on knowledge workers have never felt so obvious.
Generative AI is not some distant future. It’s going to completely throttle knowledge workers in copywriting, marketing, customer support and software engineering.
These are obvious, but surely it will upend just about any profession? Basically, any process which is logical has the potential to be composable in a model and applied in multiple ways as logic that learns.
I think in aggregate, however, ChatGPT (and other forms of generative AI) has a net positive effect on society. Sure, there will be disgruntled people who’ll be forced to overcome their mediocrity, but they’ll eventually adapt (you hope).
It’s like any historical wave of tech innovation we’ve had — the way we work will change and humans will naturally be forced to do more “creative” work.
But bear in mind, “creativity” will have little to do with execution. “Creativity” will be purely about ideas — ideas will become a greater source of competitive advantage than the speed and quality of execution (which is now openly accessible and cheap).
For example, it won’t matter how well you can read, research, draw (check out DALL-E), write, paint, sing etc — the execution component of these creative pursuits is typically logical and therefore can modelled and built upon through AI.
It will matter how well we can conjure imaginative ideas and use them as prompts to enable AI to do it’s thing.
Clearly, that’s a worrying thought for people who don’t like to think for themselves (ahem, everyone entrenched in the cogs of educational and corporate institutions).
For humans, it means we all need to get a personality as a matter of urgency. But not just some generic personality we plaster on Instagram — we need something that’s distinct and uniquely weird. Distinct personalities are difficult to replicate. Like there is only one Joe Rogan, there is (I hope) only one of me.
In the same way, humanistic values will become far more important. Some might call it “soft skills”, but I think that term diminishes the significance of values like imagination, generosity and integrity.
Neither personality nor humanistic values can be commoditised. The world needs more of it.
New models of leverage
There’s no doubt that labour leverage has diminishing value. The most viable sources of leverage come from products that can scale with no marginal cost of replication. This principle in effect has led to the proliferation of SaaS businesses in the last decade.
But even these SaaS companies have been cutting excess “fat” from non-tech functions. In the last 6 months, we’ve seen the tech bubble burst (it will hit Australia eventually). Entrepreneurs & VCs who were riding the wave of cheap money (due to low interest rates) and investing in labour to drive “growth”, are seeing the impacts of a tightening economy.
Capital is more scarce and the cost of acquiring customers is increasing, which means non-essential functions (like marketing, sales and ops) are being cut. Elon saw the writing on the wall and did exactly that. Labour does not equal scale.
This leads me to the impacts of AI on the SaaS model itself. With the maturity of AI and their underpinning models, I think SaaS will eventually be “eaten up”. Remember that famous article Marc Andreessen wrote about software eating the world, 10 years ago? Yeah, well I think traditional SaaS as we know it will also eventually die.
Software can easily be replicated — more quickly, more cheaply and without permission.
Models and algorithms — not so much. Especially if they’re underpinned by propriety data that’s unique to that individual or business.
Even with all the new AI tools spruking up in the market (like Jasper AI and Bearly AI, which are SO cool btw), they’re all effectively underpinned by the same AI data sets. So eventually, everything will end up sounding the same, because they’re piggy-backing off the same open source models.
So new models of leverage, and therefore the new models of business success, will have less to do with labour and less to do with products which have no marginal cost of replication.
It will have everything to do with products and services that employ propriety data and specific knowledge.
Naval wrote a lot about specific knowledge, and goes back to my earlier point on cultivating unique personalities.
Those that employ specific knowledge — through propriety data and a unique personality/ combination of skills— will beat the market.
Building for novel use cases
In isolation, AI bots and productivity tools are a little bit gimmicky. They blow your mind for about an hour of asking stupid questions, before we quickly get bored and discover it’s inherent limitations.
No doubt, this space will still generate a lot of investment and fun use cases — and eventually the hype will lead to an “AI bubble” in 2023. The same thing happened in crypto and NFTs in 2021 with over-priced JPEGs and fake coins.
And just like crypto, the AI bubble will eventually crash. People hoping to make a quick buck will scatter, leaving the optimistic contrarians to advocate and work on the tough problems.
Ultimately, all tech (including AI and blockchain/ web 3) is only as useful as its use cases. Right now, we’re having fun with the basic applications of AI — but the real winners in this field will use imagination to conjure new applications of this tech to solve complex problems.
The design of new solutions and new businesses will stem from the outcome we’re trying to achieve, not the tools or models themselves. And I don’t necessarily think we have to be technologists to do that — but we need to notice and care about solving big problems 5, 10 and 50 years from now.
I’m bullish on the underlying tech and it’s prospects over 10–15 years — particularly in the space of clean energy generation, re-wiring of traditionally centralised bodies (i.e. governments and financial institutions), and the proliferation of the arts (i.e. podcasting and blogging).
All in all, I only wrote this post to remind myself (more than to you), that ChatGPT is not only a necessary step forward in tech, it’s a reminder to keep writing, dreaming and wandering. That’s my highest calling as a human, and these snazzy AI tools will only help me become more of myself. I only hope they do the same for you.
You can subscribe to more random musings on tech, creativity, economics and other random shit in my (almost) daily blog.