Last month, I wrote an article predicting the end of SaaS, and people lost their minds. Well, people who didn't read the article lost their minds. Or those people who have a stake in the SaaS status quo, yeah, they lost their minds.
A couple of days ago, I asked if A.I. was just another big investor head fake, like NFTs. Not as many people lost their minds. Apparently, people love SaaS but are skeptical of AI. Me too.
I've got a long history with Generative A.I. I was there at the beginning, co-inventing the first Gen-A.I. platform used in mass market applications with Automated Insights. We exited to a PE firm in 2015, and after that, I believe they went the wrong way with the tech and I left in 2018.
The thing is, I've been skeptical of Generative A.I. for over a year now because I already know where this is headed. I know the right way to use the tech. It's killing me.
And it has to do with the end of SaaS.
I Didn't Say SaaS Was Dead
I said that the end of what we're calling SaaS is closer than you think. It's right there in the title, haters.
TL;DR; Just-fired-off-an-angry-comment: Oracle and SAP aren't going bankrupt tomorrow. But we're quickly approaching the inflection point where the user expectations and the tech are converging on a Do-It-for-Me brand of data analysis that's going to eclipse the current Do-It-Yourself model.
Thus, developers of software and apps will no longer be able to get away with just barfing out aggregated and summarized data back to the user for them to make their own business decisions. When people talk about the uselessness of apps like Google Analytics 4 (GA4) -- which is actually very useful for people who know how to crunch the data -- this is what they're talking about.
As I said in the article, the effect of Generative A.I. is going to bring about a rise in demand from customers who want solid decision options laid out for them, and who do not want to wade through screens of pretty dashboards and have to draw their own conclusions.
This is already happening.
But I don't know, some people thought I was calling for the rise of voice input over keyboards. You can't educate all the people all the time.
I Didn't Say A.I. Was a Big Investor Headfake
I just asked the question. I poked the bear. I rattled the cage.
I noted that the more recent sketchy business models and half-baked applications that I'm starting to see for Generative A.I. look a lot like the fly-by-night, money-driven schemes that poisoned the NFT/crypto well.
And if I don't ask the speculative questions, who will? Because I have the background in A.I. and also over 25 years of delivering useful products in an ever-changing technological landscape. I'm not shaking my fist at clouds. I'm connecting the dots of experience.
Here are my conclusions about Generative A.I., to which that experience has led me.
The "A.I." part of Generative A.I. is the shedload of algorithms that can now be quickly applied to the mountains of data that servers can generate and aggregate. It's decision engines at scale. This is what we were doing back in 2010. We broke AWS at the time. But in 2023? Not new. Not revolutionary.
The "Generative" part is the advancement in tech that can render the results of said algorithms on said data as a sentence, or a picture, or music.
That's cool. That's why Gen-A.I. is novel. I wish we had had that in 2010. There are thousands of fun little consumer uses and another ton of potentially scary unethical uses for tech like that. But we were approaching this tech and talking about A.I. ethics back in 2010. Not new. Not revolutionary.
But when the end of SaaS happens...
Flip the Script
I'm on the board of a new company that could potentially do something revolutionary and maybe even knock out software like GA4. These are brilliant folks. And they're taking a very cautious approach to the integration of Generative A.I. into their model.
Last week, we were looking at wireframes for their MVP, which should be finished very early next year. When they got to their "dashboards" area, the lightning struck in my brain.
They had pulled in a mountain of data, did some amazing algorithmic analysis, and put the results into a series of charts and tables that were gorgeous to look at and easy to understand. The user could even click through to recommendations for each set of data if those recommendations existed.
"Wait. Flip the script," I said.
Start with the recommendations. Spend all your time working on that. It's a big change and a scary proposition because you're doing all this algorithmic work for the user and then potentially presenting them with a blank screen of results.
Nothing to See Here. Go Back to Work
But that's no different than what SaaS does, except it makes the user do it. It forces the user to draw their own conclusions and make their own decisions. And if SaaS comes up empty on the decision analysis front, the user winds up paying people a lot of money to either tell them that they've got nothing or to make something up to justify their data-scientist-level salary.
The company I'm working with didn't start their company to be a SaaS also-ran. Their mission is to deliver value, not 1990s-style reports. That's their Generative A.I. use case. And as their A.I. gets better, they can not only tell the user what they should do but also offer to do it for them.
That, in a nutshell, is how we should be thinking about and using Generative A.I. Fill in the holes, eliminate the guesswork, and give the user confidence about their choices and actions. We've been saying this is what we want our software to do for the customer for years.
No comments:
Post a Comment