Strategic advisor to 36 practical SaaS founders | 30-year successful software veteran | Founder, Practical Founders | Host of Practical Founders Podcast | LinkedIn Top Voice
I spent five days in the SF Bay Area with B2B SaaS founders, funders, and senior software engineers. The talk is all about adding GenAI and LLM features to software apps, but they were unclear exactly how or when the AI big wave revolution will occur. AI is very cool and will be big, but it's still early to see how this will play out in the next year on the front lines of business. There's a lot of hurrying up to say you have AI features with a healthy skepticism about the user benefit of early GenAI attempts. Coders in big tech say, "Our product leaders told us we need to add GenAI features, but it still doesn't seem that interesting for users. We'll design it, code it, and ship it anyway." Founders in startup SaaS say, "We are running a lot of experiments and keeping up with the big AI platform changes. We're all more productive now but it hasn't made our app much more useful." Smaller SaaS funders say, "It's hard to tell where the winners will be to make our bets. AI tools help software companies grow more efficiently, so they don't need our funding as much. We're being very careful." Big enterprise SaaS buyers say, "We need to start investing, testing, and trying GenAI technology, which means we need to spend less on our existing SaaS apps in our tech budget." Big investors say, "It will be the next big tech wave, so we are making big early bets on chips, infrastructure, foundational models, and platforms. NVIDIA, Microsoft, Google--the big will get bigger." It's all very exciting, but it's still early and moving around quickly. That's a lot different from the last big software tech wave when SaaS apps on the web and mobile took over the last 15 years. It was easier to bet on the massive transition from on-premise Windows software to cloud-based web and mobile apps for businesses everywhere and billions of consumers. Despite that, I'm more excited about GenAI and other ML platforms than I was about VR, blockchain, and other platform opportunities that didn't materialize. It's early, but the practical AI wave will happen in ways we can't imagine or expect. What are you seeing in the current AI wave? #practicalfounders
Matt S.
8mo
The current genAI/LLM hype cycle is boring. This mindset of everything has to include LLMs is caused by a lack of creativity that stems from a lack of knowledge and experience in real industry. There are people like me working on novel AI technologies that have real world applications. These insights have been developed through years of experience within specific industry segments. And yet to get even a modest amount of startup capital is seemingly impossible. Oh well, back to work.
Saeed Ahmad
8mo
Currently majority is because of FOMO. But a good proportion is making very useful integrations using GenAI/LLMs. The trend is certainly changed, small entrepreneurs are highly into the hype and they want to build things quickly on top of LLMs but some of them are on a mission because they are crafting excellent user experiences with some custom features and decent fine-tuning of famous LLMs. I jumped into the integrations 2 months back and majority of the inbound leads are flowing into LLM integrations to automate different workflows and most of the ideas are full of potential for thier ongoing business. Well in long run, building on top of well known LLMs will be a certain win as they are going to become interface to build anything consuming their information. Currently OpenAI seems to be moving much faster than others. Let’s see how it goes further.
Kashif Sami
8mo
This is so different from the SaaS wave. While no one could have predicted how big SaaS was to become 15 years ago, the value prop and the means to production were quite clear in the early days. That's not true of GenAI. Not to forget the inherent uncertainty in its deliverables. As for LLMs in general, having been close to scores of LLM implementations in the last AI hype cycle, I walked away pretty disillusioned with the space as it got nowhere for the majority of enterprise attempts at putting them to use. I expect incremental efficiency improvements but otherwise I'm not as hopeful as you are.
Mmm...I am seeing everyone scared but a few. Next tech -call it whatever- could be the leveling field if the little guys can get smart enough to disrupt old concepts deeply ingrained in our brains. If ideas are cheaper and easier to develop, then we won't need much resources, money or time and "we" could easily compete with the biggest guys. I disagree with the big investors concept of the big getting bigger...if we do this right, the big guys will eventually get smaller and the little guys will get bigger. It will require a totally new and disruptive thinking the big guys can't do....just like Nokia couldn't and didn't.
Mitchell Tan 📮
8mo
interesting, why do you think it was different with on-prem -> cloud? was there also similar healthy skepticism around that early on?
In our positioning work, we’ve seen a lot of ai startups struggle to cleary define their target customer — another signal that we don’t really know the best use cases for AI yet. In my opinion, the early winners will be the ai startups that verticalize their approach (ie. Building their products around highly specific use cases for a unique industry) Nice post, Greg.
Chris Augur
8mo
Greg Head as usual you have cut through the noise to get at the fundamental issues. My work in Silicon Valley reaffirms the comments you shared. It will be a messy patchwork of AI implementation save for the few large LLM players. Enterprise adoption is the next wave, that's where the opportunity lies for all but the largest AI companies. Access to essential and meaningful data is a key criterion for a defensible offering or position. One must have a moat in this category.
Mark Donnigan
8mo
It's interesting how companies feel the need to integrate GenAI features, despite unclear user benefits. Tech adoption requires a clear value proposition, even for early adopters. Clayton Christensen's "Innovator's Dilemma" suggests that true disruption starts small and often is unnoticed. Companies should focus on incremental improvements and specific use-cases that solve real problems for the user. I believe that this approach is more practical than betting on big, generalized AI trends. Exploring niche applications can lead to broader success as it resonates more with early users, fostering the required traction to "cross the chasm" into the mainstream.
Founder CoRoster 🗓️ and Launchkey Lab 🚀 - CoFounder I'm in Health 🩺
8moWhat's scary is no one is really talking about how they protect their data with a super helpful AI linked to all their information on the back end and a nice publically accessible interface available for anyone to use out front. I agree it's very cool and will be big but it feels just like the mid 90's when everyone was in a tech land grab - building anything they could think of and waiting for someone to find a use for their stuff.