AI: Wait for It

...mainstream tech apps at scale take years

A common refrain since OpenAI’s ChatGPT took the world by AI Tech Wave storm in November 2022, is ‘where are all the ground-breaking mainstream AI apps and services?’ It’s a question I’ve addressed before in this tech wave since the days of the PC and Internet tech waves.

It was put in a pithy form a few days ago on X/Twitter by Ram Ahluwalia, the CEO of Lumida Wealth (@ramahluwalia, h/t @howardlindzon):

“In the Dot Com era, we had genuine killer apps:”

- “Email (Hotmail com)”

- “E-Commerce (Amazon) - The Browser (Netscape Navigator)”

- “Online communities (AOL, Prodigy)”

- “Direct messaging (ICQ, AIM)”

- “Digital Music (Win Amp, Ipod, Napster)”

- “Chat (AOL) “

- “Internet Gaming (Quake) “

“Take stock of that... Any 3 of those would be remarkable.”

“The late 90s was truly an incredible time to be alive.“

“2024 is not shaping up in the same way.”

“GPT is great for drafting, summarizing and research... but these are not at 'killer app' level. I can do fine without GPT. It's not indispensable. I cannot rely on it for investment advice or decision making.”

Those words hit home for me in particular. I was at the big table for each of the above events in the 1990s, as head of Goldman Sachs’ Internet Research Group.

Everything from AOL and its mainstream successes (not to mention Prodigy and CompuServe), to Netscape, to the online music revolution of Napster and beyond, and of course Gaming. Knew and worked with most of the principals in person.

It was the most intense period of my professional life from 1993 to 2002. Nine chronological years that in ‘Internet Time’, 7 for 1, felt like over sixty. Have many subsequent decades of closely following, investing and advising in technology waves since then.

Here’s the Internet Tech Stack as it turned out back then. Again, worked closely with most of the people at these and many other companies in that frenetic period:

None of the things Ram highlights in his tweet above happened in a year. It all took the better part of a decade, with a whole lot of financial ups and downs.

AOL in particular took over seven years to get to ‘Scale’ with over 25 million subscribers, and wield a market cap big enough to briefly buy the biggest traditional media company of the time, Time Warner

Technology takes time. At least a few years.

It’s been barely a year since ChatGPT was this AI era’s ‘Netscape moment’ of 1995.

Give it time. At least three years. Probably five.

2024 is just the year it barely gets going this early in the AI Tech Wave chart below.

But likely not as long as the decade and more we’ve waited for other tech waves in recent memory like ‘self-driving cars’, crypto, AR/VR, and so many other eagerly awaited technology change agents. They’ve all taken over a decade and a half and tens of billions in public and private investments.

Think we’re going to be startled positively by the exponential mainstream capabilities of Foundation LLM AI technologies just in the next three years in both consumer and business applications. I’ve written a lot about many of the possible ways in these pages. Daily now for over 250 days. Including the financial forces behind it all.

The other thing to highlight is the ‘Small AI’ part of this AI Tech Wave has also barely begun. Most of the focus over the last couple of years has been the ‘Big AI’ Foundation LLM AI model driven set of innovations. Bigger and faster AI software and and hardware GPU driven data centers, doing exponentially more impressive probabilistic ‘reasoning’ driven tasks in inference loop driven software.

It’s like we’re seeing the mainframe/minicomputer part of the computing revolution which took decades before the PC/Internet wave started in the 1980s and 1990s. This time it’s all happening almost together, barely a few months apart. 2024 is barely the year Small AI starts as I’ve noted many times before.

So let’s wait for it all, big and small, just a bit longer. It likely won’t take too long. But it most likely won’t all happen just this year. Stay tuned.

(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)