Are We Effed?
Brent (our founder) recently sent me a Twitter thread from someone who declared that we are effed. He knew a software developer in Pakistan who used ChatGPT 4 to study a hundred thousand Shopify sellers and then email three thousand of them each a plan for a unique custom app that it ideated that the software firm might build for them to improve their business. It was a massive, audacious cold email campaign and apparently it worked. The response rate was off the charts and in order to handle the volume, he also had ChatGPT 4 respond to the responses.
Minds were blown… The AI learned about the target company, learned about the capabilities of the software firm, brainstormed reasonably good ideas at scale, and then credibly interacted with other humans on specific topics.
After he sent me the thread Brent wanted to know (1) how we might employ similar capabilities to benefit our business and (2) what the endgame of widespread adoption of this technology might be.
With regards to (1) the answer is really any detail-oriented, repetitive task. For example, ChatGPT could be trained to look at company financials, identify red flags, and engage in initial due diligence over email, handling both initial questions and follow-ups, particularly if you helped it figure out what to look for. Now, you probably wouldn’t want it doing these things unsupervised, but if it could get you 90% of the way there, that would result in a significant efficiency level up.
Which brings us to (2)…
If you work online, you are about to get flooded with AI-generated solicitations. In fact, you probably already are (and Elon Musk already admitted as such). In fact, I kinda sorta suspect based on the tone that the original Twitter thread Brent sent me was generated by AI as well (and how meta is that?).
Being online is already a terrible place to be. Studies show that it’s bad for mental health and that it’s contributing to polarization. Moreover, my own experience is that the signal to noise ratio is deteriorating.
So what’s it going to be like when infinite bots are generating infinite content?
This is the dystopian turn our conversation took, but it ended on an optimistic note. In this world that may be nigh, the value of two very human qualities should skyrocket: judgment and authenticity. Judgment because in a world of infinite content it is going to be very valuable to be able to differentiate the good from the bad and the high potential from the low. Authenticity because in a world of infinite content the only content that anyone will respond to will be words that really resonate.
What’s more, are we going to become cynical and skeptical about who we think we’re interacting with? If you think it’s a bot on the other side, how likely might you be to engage? If very, what are you going to do?
We thought about that and wondered if it might drive people to deepen human, in-person relationships. You’ll reach out to the people you know and want to talk with them directly. Workers are already returning to the office in droves following a few years of unsatisfying remote work that many initially thought would be a boon for productivity and finding that they like it. It turns out we’re social creatures.
So whatever else happens with AI, if it puts us back in real world contact with one another, that would be a real benefit. And an improvement. And perhaps help us not be effed. Because judgment and authenticity matter.
P.S. This was written by ChatGPT 4.
P.P.S. Just kidding…but how might you tell?
– By Tim Hanson