There's a Chat For That
Consumer and Enterprise LLM Ecosystems Are Changing How We Compute
First, a little personal news (related to this topic) and why I’ve been light on posting recently. I just finished my first week at Ringer Sciences, where I joined as EVP of AI Analytics Strategy. You can read more about that here. I’m more than excited to take another step forward in working in AI since I immersed myself in the Conversational AI space over at Soul Machines with interactive AI avatars.
Now, let’s get to the business at hand—with a story from the past to help us peer into the future. Before mobile apps, an actual mobile Web, and advanced smartphones, there were Blackberries and early generations of smartphones with tiny Keyboards and screens that gave people like me the smallest sneak peeks into what was coming next. Then came the first authentic glimpse of what the modern mobile experience was to become—the iPhone. I’ll be honest here (embarrassingly): I didn’t get it at first. I LOVED my tiny little keyboard on my (ahem, Sidekick) and balked at removing it.
But I believe in Apple products—all of that changed about five minutes after getting the first generation iPhone. And so, I was one of the first early adopters to have one, and I have never looked back. I learned from that experience, and so when Apple first announced the App Store—I immediately understood what a game changer it would be, and it exceeded all expectations. Apple’s App Store has created a robust ecosystem where thousands of developers create massive value for Apple and its users of Apple products—it has influenced scores of business models to pursue similar strategies. The Apple ecosystem works something like this:
-Create, own, and tightly manage the ecosystem
-Harness the unlimited amount of “free labor” that goes into apps
-Take a considerable slice of the money apps make
-Make even more through “free” models through an advertising network
The Ecosystem Is The Product
Open AI’s recent deployment of GPTs that anyone can create is a big deal and, like Apple, is already being described as an “App store for AI,” which is catchy, though it’s not entirely accurate. All kinds of AI applications do very different things, and it’s yet to be seen if ChatGPT on the consumer side will do all of them better (text to video, for example—Runway is best in class until Adobe either buys it or makes a better version). That said, Open AI and ChatGPT have built the foundation for a consumer version of an LLM ecosystem, which puts the power of LLM development into the hands of consumers, developers, and creators alike.
GPT creator allows both the average person and a developer to create a custom GPT-powered LLM, training it by uploading files and/or pointing to the Web and even importing APIs. I spent a little time creating one (it needs more work), but I also tried out one of the initially Open AI-built GPTs. called “Data Analysis,” which is fitting given my new position. It efficiently analyzed the spreadsheet that I uploaded and gave me several options to visualize the data:
With GPT creator in the hands of millions—we’re going to see the invention and creation of all GPT-powered experiences, and this is going to do three things:
1. Create free labor for Open AI’s “App Store”
2. Pour terabytes of data into Open AI’s servers
3. Accelerate conversational computing (interfacing with technology by talking to it conversationally)
There’s A Chat For That
It’s the last point I want to dig into more, as I’ll leave it to others to analyze the tech sector. I’ve already written about how every business will have an LLM if not several. The consumerization of AI powered chat-based experiences that actually work, thanks to advancements in LLM, will only move faster, both on the consumer front and the enterprise. The problem LLMs can and will solve is that we’ve reached a data and information saturation point that’s become unmanageable. We sift through e-mails, intranets, servers, Websites, blog posts, more servers, etc., to find the information we’re looking for—and even when we do—we don’t even know if it is accurate or if it fits the context of which we need it.
LLMs, much like the ones we are seeing for consumer use (and will see more of), can help solve this problem at the business level. Prompting in a chat interface or conversing with an avatar intelligently and providing the information (and/or access to source files) is a superior employee experience. Enterprise LLMs possess the potential to help teams do their jobs better, faster, more effectively, more efficiently, and more accurately—acting as essentially workplace assistants that begin to bring back together all of the fragmented, siloed, and potentially outdated data.
The Data Mining of Prompt Fields
Lastly, all of the above—both Open AI’s ecosystems as well as what will be built on the enterprise side of things are already making an empire of data that is going to rival what we’ve seen created by companies like Google, who sit on treasure troves of insight because they know what people search for, how, when, where, and can translate what it means.
The prompt field is the new search box.
This is the paradigm shift that Google is all over right now. They understand that prompting, the things we are asking LLMs to do for us, shows intent and gets into even more nuance that searches can provide (as rich as that data is). Prompt data will create an entirely new way to look at what data is and what it means. The power will go to the platforms sitting on all that data—this is one of the reasons why companies are being selective about how they go about building their proprietary LLMs—who can be trusted both with the data that the LLM trains on, what it can ingest from users as well as what the prompting from employees reveals.
In short—all of the above adds up to an “app store moment.” The power of LLMs is in the hands of the consumer, and so is building one. Inevitably, all of this leads to the “consumerism of IT,” which happens today much faster than it did ten years ago. So let’s say you work in the communications department of a large company and want to draft a communication reflecting the latest. Sustainability efforts ladder up to the company’s mission, and you need help putting it together…
There’s a chat for that.