Feast or Famine

Since November of 2022, AI has become everything, everywhere. I have followed and ridden cutting-edge tech waves for years. I learned the hard and valuable lesson in 2008 of the need to always stay current.

In 2008, I was running a software consulting business in Boston. I had a strong focus on e-commerce. This was back when we built bespoke e-commerce solutions. We did very well during the early rush for businesses to get online commerce into their sales channels.

The 2008 financial crisis was a killer. Overnight people put their credit cards away. For the next two years e-commerce sales did a nose dive, and so did my approach to getting business online.

I took a few important lessons from that period. These have stayed with me, and I have forgotten, excused or ignored these at my own peril. Someday soon I will unpack these and see how they impacted my last startup (Meenta.io R.I.P 2022)

Learning #1: Speed Before Beauty

The US market is too big for single-winner solutions. Any tech solution, pattern or service has both visible and invisible replacement options. Any good idea is not unique. Someone smarter and better-looking is already working on it.

Get to market fast! One or two good features can give you the market. Waiting for perfection is a bad idea. Dirty solutions are fine.

Learning #2: All Tech is Cumulative

You can not pick and choose which innovations you decide to adopt! Every innovation adds some level of productivity gain. The gains might be measured in tiny increments, 0.01% more productive. But with time these micro productivity gains add up. Skip tech innovation and you lose the learning lessons that allow for the addition.

Skip a few innovation cycles and all hell brakes loose.

Learning #4: Crowd Source All Tech Questions

Open Source is crowd-sourcing. There are plenty of reasons to be careful with it. But overall it’s the source of rapid and sustained innovation. Period. In 2009 I was still manually resizing images, versus using OS PHP image libraries. Fail.

There are just too many options for any one person to internalize. Even if I was funded and could attend a new tech conference a week, I could never stay up on everything, let alone process all the techie goodness.

My approach is go to my network! Got a question? ask my network. If I don’t have someone to help, then I build the network. Making this pattern my default has unblocked a ton of solutions.

My Network Learning for AI

Unless you are a CEO+CTO wunderkin, the AI craziness of the last 24 months can be like watching the technology equivalent of the Big Bang in real-time. A lot of energy, new companies popping up like new stars, and a stupid fast expansion. This speed leaves little time to consolidate patterns.

The first step in wrangling all these signals into a strategy is to get a sense of the players. Here is my AirTable AI collection. This grows every week, as my network tips me off. Google is just not good enough when things move this fast.

TLDR

Building an AI strategy using off-the-shelf tools is a fast way to gain the benefits of this proliferation of tooling, without the massive overhead that adding fully backed AI can provide over time. These tools can be the AI equivalents of No-Code components.

Post Big Bang?

As with the actual Big Bang, the initial expansion will be followed by a dieback. The same pattern occurred with the advent of new databases that evolved to meet the noSQL and mobile ecosystems in the early 2000s. A number of these new AI tools and services will not make the revenue cut, and winners will emerge.

But right now, we are still in the early expansion period. When I got the first whiffs of the changes in 2022, around the time FTX was starting its meltdown, I went to my network.