Riding the AI Wave: From Fart Apps to Fortune

Exploring the potential of AI’s paradigm shift, the challenges ahead, and the massive disruption it’s poised to unleash.

When I first saw ChatGPT last year I thought, “Oh, wow, somebody solved AI. Cool. We can get down to the business of watching OpenAI turn into the next big super-multi-mega-corn and figuring out where to put the “O” in “FAAMG” (“AA OMFG” gets my vote for obvious, childish reasons). As I started playing with the open source models and seeing the Google leaked document about the playing field and pace of innovation a light bulb went off.

I took the leap and decided to start playing a little more heavily with AI. ChatGPT is super interesting along with many of its other contemporaries in the space; Midjourney, Stability.ai and a near infinite number of really compelling open source models. Like others, I’m seeing this for what it is; a tectonic paradigm shift on how we’ll work in the future and it’s impact is going to be felt far wider and faster than any other technology before it.

Back when the iPhone hit and “apps” were all the rage, I got laughed out of more than a few partner meetings while trying to raise money. I seriously got asked, “But how many fart apps can there really be in the world?”

I would put AI in the “fart app phase” of it’s existence today. People playing with it, finding interesting solutions that scratch the surface but only starting to see where this is going to go. Just like when people had to have a website during web1 and you had to have an app when mobile hit, the same is true of every company right now touting their AI-this-and-that in the media and on their websites. The real magic happens when these solutions take hold and fundamentally change business process. Getting something delivered from a brick-and-mortar retailer via their website or getting to order ahead through a mobile app for your coffee; these things don’t happen overnight and there is a set of organizational and human infrastructure that must be lifted into place to make them happen.

The same is true for AI.

AI is going to manifest itself as a universe of finely tuned large language models that interoperate and change how business and human/brand interaction is done.

So many questions come to mind:

• How do we securely and safely train these models with “ground truth” data?
• How can we sidecar data into an LLM and do so in a protected manner?
• Privacy and PII controls? You can’t just shove a bunch of sensitive data at the mega-clound and hope you won’t have spillage or people who have access that shouldn’t.
• How will we train and manage LLMs in real-time? When you train it up on data today, how do you continue to iterate on it and manage the output over time? (think of the legal use case here)
• How will LLMs interoperate?
• Devops?! Who the hell is going to manage all the devops?!
• How will LLMs be safeguarded from a security perspective?
• Inevitably, as with web and social, AI will be weaponized, but how?
• What about an LLM catered to every human on the planet? Your email, calendar, browsing history, likes, watches … all of it fed into a bot that you can use to find the next thing or something you can’t quite remember.

That list alone there has to be at least $5 Trillion worth of disruption and business to be had if not significantly more.

AND WE HAVEN’T EVEN SEEN THE AI-specific HARDWARE YET. That is a good 2 years away with pretty much all of FAAMG on some iteration of it.

Hold on tight people. This is going to be one wild and bumpy ride.

I cannot wait.

Riding the AI Wave: From Fart Apps to Fortune was originally published in 𝐀𝐈 𝐦𝐨𝐧𝐤𝐬.𝐢𝐨 on Medium, where people are continuing the conversation by highlighting and responding to this story.