Convictions and principles in epistemically indeterminate conditions

From ChatGPT Plus is here — and it has Google in its sights by Malcolm McMillan.  

This news comes on the heels of Gmail creator Paul Buchheit tweeting, “Google may be only a year or two away from total disruption. AI will eliminate the Search Engine Result Page, which is where they make most of their money.” Given that OpenAI also announced plans for a ChatGPT API waitlist), lower-cost plans, business plans, and data packs, it’s clear that the software developer has its sights set on ChatGPT becoming a go-to resource. 

Nearly a decade ago, I had a conversation with colleagues in Silicon Valley.  There was grave concern about increasing tech company vulnerability to excess and misguided state and federal regulations. 

Our discussion was robust without being antagonistic.

The argument I made was that we were seeing the real world collision of competing models.  There are those who view the world simplistically as a process.  For any problem there is a process solution.  Which often takes the form of regulations.

Then there are those who view the world as a rich and disconcerting aggregation of intersecting and subsuming systems.  A change at point X at best has a probability of driving a desired change at Point Y.  And it may be a small probability.  

You can characterize the contrast as that between Utopianists and Pragmatists, Deterministic Engineers versus Systems Thinkers, Regulating Government versus Free Market, etc.  

My point was that Google's dominance at that time (and still) of the Search market was hugely profitable to them but also a vulnerability.  For all their dramatic and interesting investments across a huge spectrum of ventures and fields, none of them were generating more than 5% of revenue.  All the profit and most the revenues came from Search.

So there was an acknowledgement of the two-edged sword.  Their natural dominance of Search was potentially a regulatory problem generating discussions about more control or even of breaking up Google.

Viewed solely from the Utopianist worldview, there was a clear problem (dominance) and a clear solution (regulation or breaking up).

My alternative, the market alternative, was that all dominant positions eventually crumble.  Especially dominant positions in dynamic sectors.  I argued that nothing needed to be done because 1) we did not understand the problem (dynamic chaotic system) and because 2) the cost and consequences of the preferred regulations could not be calculated or confirmed to be net beneficial.  

All we needed to do was ensure that there was as little abuse of dominant position as possible and then wait.  Something or someone would come along and undermine their position.

To which the Utopianist response was naturally:  What competitor, what competing product.

"I don't know" is the only correct Market response.  Because we don't know till it happens.  

We know the market is competitive, dynamic and evolving.  We know from history that this always eventually unseats the dominant with some new competitor taking their place.  All you have to do is look at the constituent companies of the S&P over time.  Giants come and go.

"I don't know" was not a convincing argument with the Utopianists with their process engineer view.  

It is an age old contest between principles and broad rules in general systems thinking versus deterministic tactical solutions in the Utopian world.  

A decade ago I had no vision about chatGPT as a mortal danger to the Google business model.  I certainly don't have a particularly informed opinion at the moment.  And maybe AI isn't the threat to Google that is being alleged.  But if it is not AI, it will be something.  The competitive market is always seeking advantage.