David Evans’ Post

Vertical AI may or may not be the future of AI, but it's certainly the current cycle in AI. Controlling outputs from AI applications is much like a wooden labyrinth game. You know the path, but keeping the ball on the path isn't quite so straightforward. The smaller the game board, the easier it is to navigate the maze. That's the push for vertical AI. By controlling the range variables, we effectively keep the data game board smaller and easier to manage. Plus, by going vertical, it allows for getting deeper insights. The challenge long term will be scale. Most vertical organizations will be pushed to flatten to grab larger and larger market share as they scale. So don't be surprised when Bessemer starts talking about flattening the curve. https://lnkd.in/gcHeeuwU

Before starting to examine the strategic questions that an acquirer will ask, you need to understand whether you are looking at a financial or strategic buyer because the goals will be different. A financial buyer is looking at the cash flow of the business. They're looking to generate a return from the investment based on the profit and associate growth. Financial buyers will care more about growth metrics, cost of goods, and overall profitability, i.e. objective measures. A strategic buyer is looking at how it fits into their operation. While they may still care about financial metrics, the real upside for them is based on how it helps them sell their core products and services and even help that growth. Strategic buyers will care more about how complementary you are, i.e. subjective measures. The reason it's important to understand the two is that the former is likely willing to pay less to acquire a company than the latter. https://lnkd.in/ghfwB4de

Fine-tuned retrieval augmented generative systems (RAG) from companies like LexisNexis still hallucinate, or generate errors, in about 17% of cases. That's a far cry better than the 40%+ of generic use models like GPT, but still not good enough. AI does not deal well with subjectivity because, after all, at its core are still mathematical models. And at its core, the law is subjective. But there are a lot of things that are subjective, from medicine to business strategy. The key mental model is to stop trying to build systems that have the answers. Build systems that let users comb through more data and synthesize that into rational responses. https://lnkd.in/gQKsbT7R

Thin applications that solve the limitations of other software systems have been around forever. Things like antivirus or web browsers were all once unbundled, and in SaaS, everyone tries to make systems like Salesforce not quite so hard to use. AI has exacerbated this problem because for most of the last two years, AI-driven features just didn't exist in the underlying product. The challenge with "thin wrapper" products is that they are not truly innovative. Real innovation means you need to re-think the entire process, the data, and the outputs. If you don't, you're just one product management meeting away from becoming a part of the system you sit on top of. https://lnkd.in/gmkNuSna

When talking about AI, I'll generally use a broad umbrella that includes both generative AI (ex. ChatGPT) and machine learning (ex. predicting when a machine will fail based on historical data). In both cases, the machine is using data that it has seen in the past to extrapolate a future state. With machine learning, that prediction is generally a finite result: predicting a numerical value or the most likely option. So the outcomes have some level of predictability. With generative AI, it's a more amorphous result. Using a batch of text or image input to produce a variable output of text. Because of the breadth of inputs and outputs, generative AI can be less predictable. But the reason why they get used interchangeably at times is that, at their core, it's using predictions to determine an outcome. The only real difference is in the complexity of the process and the associated results. https://lnkd.in/g4V3jd64

Yes, churn is bad. But you also need to remember that your product, while the most important thing to you, is likely far from the number one priority of your customers. So, as much as having great onboarding and customer support functions are critical, it's also imperative to ensure that you're not becoming a chore to work with. Are you slowing down time to value with yet another training or walkthrough session? Do they really need to hear from you every 3 months with a feature update? At the end of the day, you're a piece of software that they use to complete their job. Enable that and get out of the way.

Want to know if your job is safe from AI? Ask yourself how much of your job requires deep thought. About 11/12th of a coders work is thought. I can attest to this firsthand after years of solutions devised everywhere except at my desk (in the show, in traffic, etc). While AI can generate code, it's can't replicate these processes. Thought isn't encoded in the source data for AI models. Code, documents, presentations...all of the written output we have fed to AI is a compressed version of all of the context and reasoning that went into that final work product. In other words, the models are missing 11/12th of the work. So, the aspects of your job that involve thought are safe. The rest, not so much.

I'm as much of a profitability wonk as anyone. At a high level, startups should be efficient with their cash spend, burning $1 or less for each new $1 in revenue. However, as with most KPIs, this isn't a unilateral assumption. For companies pre-Series A (and even after), there will be times where you will need to make capital investments in products and infrastructure (both human and physical). During these times, when a company is small, your burn may well exceed the magical 1X. And that's exactly why you raise venture capital: to get through the capital intensive periods of early growth in the business. But the important part is that those periods of excess burn should have an asymmetric return potential. That is the investments being made during these periods should be able to return far more than 1X in the long run.

Agile software projects, where teams that include business users and engineers iteratively work on features, recently have developed a 3X higher failure rate. But are teams really using the intended agile methodology? The core problem with the usage of agile today is that it has become an excuse to avoid focus and change plans on a semi-regular basis, because generally agile projects define priorities on anywhere from weekly to monthly timelines. But, the original precepts of agile are based on engaging the user in the design and decision-making process, not just a scheduling mechanism. While that is still partly true, most projects just run off an infinite list of things the team would like to do that may or may not have a broader correlation and vision. And that's where agile tends to fail: when teams abandon a strategic project vision for the sake of pursuing a shiny new feature because the scheduling mechanism of agile allows it. So instead of allowing agile to be a built-in distraction mechanism, use it as intended: a heartbeat to ensure ongoing alignment on the strategic project vision. https://buff.ly/3Rqb9L8

VB6 was one of the first development environments to include auto-complete. Rather than having to look up various functions, it gave you suggestions as you typed (a simple lookup, not AI). But that was a massive time savings for me as a newly minted software engineer, and it massively sped up the learning curve. We're seeing that now with AI code completion where 50% of the "typing" is done by a more sophisticated version of that early auto-complete. And we should expect a similar outcome: massive time savings and reduction of the learning curve. But let's be clear on this metric. This doesn't mean half of software development is being done by AI. It's still the human managing the requirements and broader context of the code. It's no different than cooking. Would you say that the chef isn't cooking the meal because the vegetables were pre-cut?