HERE. WE. GO.

I want to focus on B2C, originally this was going to be B2B or internal operational needs. That’s not even the exciting part. (Think about narrative gaming experiences as well)

I think that it is interesting to think about how LLMs are going to be used in the B2B context and asking the question around what are the results, the context, and the sentiment that needs to be included in the correct response. Verticalized applications need to be deployed in a hardware conscious approach. Current requirements to deploy and engage with a language model would make the Spanish Armadas’ toes curl. Something I am considering is “how can we deploy these models in a node/containerized way?” The idea of the cloud for most businesses is that consumption of hardware isn't occurring, that consumption exists in some nebulous, far-off, mythical echelon. The idea of on-prem, hybrid, and total cloud approaches are far off from most people's understanding of how an enterprise jobs execute, if they think about it at all. I think that there’s been a lot of excitement about the text and image generation aspects of generative AI and it deserves the proper respect, but if you want to blow your own mind around the possibilities of this technology think about it when it can deploy without human touch and at a scale. I am jumping into the subway scene from “Minority Report”, but only if we can figure out the personalization factors both in UX and scaled deployment. You think Tik Tok has you figured out, wait till you appear in a random subway wearing clothes you’ve never even considered yourself in. Now don’t start crying to me about data governance and owning your privacy, you will give-in because the price of convenience is going to be so great. This will have to be balanced with the projections of our economic condition but I am extremely bullish on western consumerism.

How do we get to commoditization?

First we need to consider what the characteristics of commoditization truly are. Will water, wood, or steel ever be traded on the same market as a tech company started by 4 Patagonia wearing Stanford grads (possibly PHDs, so don’t mean to offend) specializing in vertical LLM applications, I think its sane to say “No”. Luckily for people that enjoy transposing big words that most people have to google, commoditization is the correct word to consider in the great “disruption equation”. Grant me a quick aside, the condescending tone should not be mistaken as a doubtful view of what this technology is going to do or what it is going to be achieved. My point-of-view consist of a blinding hubris, that forces me to talk about LLMs with anyone that will listen. The magnitude of productivity gains for the enterprise space and overall humanity is what Balaji Srinivasan thought the blockchain was going to be six months ago.

Back to commoditization: This prediction is not going to be correct, let me start with that — but I would say in five to ten years we will see a fairly inelastic product. Since, I pulled out my $40,000 a year college words, what does that mean, the proliferation of LLMs will be so necessary in how companies conduct business internally, but more importantly how they serve their end customers. Product substitution makes me fairly quesy about this prediction. Why? Because after Facebook we saw a million network applications, its the next “thing”. In that space, generally, the value for the user is contingent on the value of the network. Generative AI will get the same kind of attention, but differentiation will be extremely difficult and my prediction is based off of vertical ownership from data input, personalization, sentiment analysis, model deployment styles et cetera et cetera. I will say that after seeing the number of milk brands that exist at the store tonight that externality has subsided ever so slightly, boy its crazy how many you have to choose from, almost overwhelming.

The Cursive Demise

Thinking about LLMs/generative AI forces my ADD to carry out stochastic terrorism on my neanderthal brain. You can probably tell if you have gotten this far, because I still haven’t really made a point. I would really like to think about how we got to where we are today. This is going to require a lot of text for me to type so hang on and I will try and make it interesting for you.

Come back in time with me to an angled roof with an endless number of short book cases in a block like formation (tried to paste a picture, didn’t work) around an endless number of tables with chairs that my left leg couldn’t fit into anymore. String theory straightened wood grained tables and the deafening sound of pencils breaking, snapping, and dragging across the pre-drawn 0.5mm “Practice Your Cursive” worksheets. For me, this is like second grade, I think. Then out of nowhere like a B-2 bomber appearing out of the sky at a college football national anthem singing, we never see one of those practice sheets ever again. Now if I remember correctly, I did not complain even one bit, until I was thrown into the baby shark inducing panic from the unison clicking and clacking of 10 year olds learning to type on the beefiest keyboards ever created. It was quite evident that my school district wasn’t in the business of buying Rams/Jobs quality products. This was the vanguard moment, when the decision to commit an entire generation (Gen Z) to the world of the computer occurred.

#The point I want to continue to write about is this: we have come into this digital age with these landmark events that have fundamentally shifted out expectations of what the future looks like. Generative AI for me is the piecing together of all these elements. After this we will go to Amazon - shopping and AWS, Facebook and communication, and then wrap into how things are changing economically but also shifting out expectation from companies around service — maybe also mention how peoples job expectations and feeling around the service industry will give this fertile landscape for generative AI to take over customer service industries.