Locator: 49314LLMS.
Locator: 49314AI.
Locator: 49314CHATBOTS.
Updates
November 2, 2025: nice investment story at this link -- Anthropic / Claude. At wiki.
Original Post
Twenty years from now, we may look back at this chart and realize where things changed.
Back in the early 80's (1980's) with regard to personal computers it was about 98% Wintel and 2% Apple. Something like that. Somewhere along the line something changed. There must have been an "a-ha" moment when some really, really smart investor realized what Steve Jobs had accomplished and what he was going to accomplish.
So, now fast forward to chatbots.
First of all the jargon.
AI
I may be wrong on this but for now this how I interpret the jargon.
AI is AI. I think we all have our own understanding of what AI is. And then there's generative and inference AI. Most of us just put all that in AI. I don't know. To me, playing chess against a computer is AI. Having the computer explain why he/she/it did what it did is generative or inference AI. But here is the real answer:
Inference and generative AI are different: inference is the process where a trained AI model uses its learned knowledge to make predictions or decisions, while generative AI is a specific type of AI that uses inference to create new content, such as text, images, or code.
Inference is a broader term that describes the use of a trained model, while generative AI is a category of AI focused on creating novel outputs.
Now, I get it. When ChatGPT asks me if I would like a graphic of what it just provided me via inference, AI will provide new content, such as a graphic which has never been produced before using generative AI.
Okay, so that is AI.
I have a tab at the top of blog, AI. That's where I update chatbots. I figured that out on my own some time ago but do not recall how that happened. So, today, to refresh my memory, I asked Google, are large language models simply AI?
The answer:
So, in the big scheme of things, that AI tab is absolutely correct.
AI needed a way to make that "information" available to humans. To do that, chatbots were invented. The chatbots were the new human interface between machine and human.
The New Human Interface
To Me As An Investor
Once, the machine spoke in numbers — a flicker of zeros and ones, cold and absolute. We built the graphical user interface to make its logic visible, to turn computation into something the hand could touch and the eye could command.
The mouse and the window were not mere tools; they were acts of translation, rendering binary thought into a landscape of icons and movement — a world of sight.
Now, the machine no longer hides behind the screen. Through the language model, it speaks — not in code, but in words that seem to think.
The chatbot is the new interface: where visual metaphor gives way to linguistic dialogue.
It is not a tool for the hand but a companion for the mind, translating probability into persuasion, data into discourse. If the GUI made the computer visible, the chatbot makes intelligence audible. It is the next interpreter between silicon and soul — the new human interface, where we no longer click or drag, but converse.
Chatbots
GUIs: look at the timeline of GUIs to get a feeling for how fast chatbots have taken off (next paragraph).
Now, compare with chatbots. The chatbot timeline. There were some ur-chatbots as early as 2010 (Apple Siri) but in the big scheme of things, it "happened" in 2022 with ChatGPT. My son-in-law, an engineer by training and an early adopter first discovered ChatGPT in late 2024 or early 2025, so I'm pegging the chatbot timeline as beginning in 2024 for the early non-technical American adopters.
The Shakeout
When looking at the timelines of both GUIs and chatbots, this stands out. In the beginning, there were many. At the end of the day:
- GUIs: two winners -- Microsoft and Apple.
- Chatbots: still developing -- but I bet we see the same thing -- perhaps three or four prominent chatbots.
Why This Matters
One graphic (note LLMs and Chatbots are used interchangeable):
The five LLMs above:
- OpenAI: has not yet gone public; has just marketed a "for-profit chatbot."
- Meta: which by the way, can in some ways be credited with the first real chatbot.
- Anthropic: huge gain in market share -- free version available.
- Google (Gemini): huge gain.
- Others:
AI prompt: There's a graphic on x today that shows OpenAI chatbot dropping from 50% in 2023 to 50% in 2025 with huge gains by Anthropic and Google. Why have Anthopic and Gemini shown such great jumps in market share compared to decrease in OpenAI's market share?
Reply:
For investors: there are ways to think about this. Maybe tomorrow. I think I know how Apple could play it. Jim Cramer says the same thing. One word: an auction. A federal judge has already telegraphed how it might be "seen" by the courts.
The Last Browser
For half a century, the web has asked us to search. We learned to navigate it through portals of glass — Chrome, Safari, Firefox, Edge — each a window opening outward to an ocean of data. The browser was humanity’s first prosthetic for curiosity: a tool not for knowing, but for finding. It required patience, literacy, and a certain faith that the right words typed into a blank bar might reveal truth from the void.But the next transformation of the web will not be navigated. It will be spoken to. The rise of the large language model marks the end of browsing as an act of exploration and the beginning of knowing as a form of dialogue. The LLM doesn’t show you where knowledge lives — it becomes the knowledge’s voice. Instead of summoning static pages, it performs a living synthesis, fusing sources, contexts, and patterns into something newly alive each time you ask.
In this light, the old browser becomes a relic — a compass in an age of autopilot. The icons at the bottom of the screen, those portals to infinite tabs, will remain a while longer, but the frontier has already shifted. The AI interface is the new medium, collapsing the gap between search and understanding.
There will always be reasons to “visit” the web — to check the shape of things, to see with one’s own eyes — but increasingly, the browser’s primary function will be absorbed into conversation. The screen that once opened onto the internet will soon become the conversation itself.
And so the age of the browser ends not with obsolescence, but with transcendence. The window closes; the dialogue begins.
