AI Eats world?
It doesn’t feel too long ago that markets were going into full on panic mode that we were at peak AI demand and spending, with the news of Deepseek. Now, the pendulum has done a full swing in the other direction and nothing is going to exist once AI has had its chance of doing the job. If you do your job through a computer, good luck! If you work in software, dust off the resume! Lawyers, Programmers, Doctors, Consultants, all on notice! Of course, I am being facetious but the general theme around AI has indeed taken an extreme shift toward this idea that if AI can do a bit of a job at this stage, it will do the full job soon enough. While this will certainly be true in certain areas, the reality will be somewhere in the middle and will likely look a lot more like something where AI tools augment people in their current roles, accelerating their output and efficiency. We could philosophize and opine for many pages on what and how AI will impact each corner of our life but the truth is no one really knows with confidence how it will unfold and evolve. What we probably do know with confidence, however, is that while it will be disruptive in some ways, it is not going to lead to some sort of business, economic or employment wasteland. Given the topic of impacts is so wide we want to drill down into the specific topic that is getting extrapolated to absurdity at the current moment which is AI and how it will eat software. Let’s look at some of the concerns being levied in this debate and see where things land.
Competition
One common concern is that ‘anyone’ can create software now. But this has always been true. Yes, it is ‘easier’ but the barriers to entry for software have always been low. It was never creating the software that acted as the moat. It was getting to scale and having the resources to crowd out (or buy) your competition before they became a threat. Moving fast in software has always been a risk and strategy in order to be able to claim first mover advantage.
In fairness though, the risks of a new company/offering coming to market at blinding speed and taking share from a competitor is higher now but this has always been a fairly high risk from day one in software. This whole topic reminds us of the famous quote from arguably the best TV show ever made, The Wire, where Slim Charles tells Cutty that the ‘Game’s the same, just got more fierce’.
In-Housing
We are still early, so we might be proven wrong here but we think the idea that all of these large companies are going to start creating their own in-house software via AI might be the most misguided one. Technical debt is real and this idea that large incumbent companies are all of a sudden going to create these fast-moving AI teams and try to break their current tech infrastructure up for some AI powered one seems again misguided at best and probably dangerous from an operations perspective at worst. Just like with software, most companies did not take on software carte blanche, instead they plugged in different services and offerings on top of what was already there. In fairness, there is an interesting world where all of these software apps get collapsed down into a centralized AI tool but those same software tools are still likely running in the background opposed to being replaced outright. So, maybe you don’t interact with them in the same way, but they are still there doing the work that needs to be done. Regardless of how this piece evolves, the idea that businesses will be creating these vertical systems and tools in-house is probably not a very realistic one in most cases.
Lower Margins
This is a bit of a chicken and egg issue. On one hand, increased ease to create ones’ own software (even if it’s just used as leverage when negotiating price points with vendors) and more demand within budgets for AI does have potential to pressure software company margins and/or demand. But what gets missed here is that those same software companies can also use AI to increase their own margins or even sell new, higher margin products that utilize AI to their customers. So, while there might be some lumpiness on how margins evolve this one seems a bit like its turtles all the way down. We are also seeing instances of companies that are utilizing AI tools to reinvigorate their growth rates. Whether it is attaching additional AI services to their current offerings or using unique AI tools as a land-and-expand strategy, AI is also bringing growth and margin enhancing opportunities to some companies.
There are also some themes that are getting missed in this whole AI eating software debate.
People like someone to blame and yell at when things go wrong
Utilizing third-party software is nice because when something breaks, you can call that company with hundreds or thousands of employees who will scramble to fix it. You also have access to niche knowledge workers that are less likely to be on staff in-house. If you create all of these tools and service instances in-house, heaven forbid something breaks or a key employee is on vacation! Companies outsourcing their non-core business needs to third parties is common and makes sense because it allows a company to do what they do best. Ironically, over time AI likely leads to further process outsourcing to new or existing companies as AI allows an increasing number of processes to be done efficiently and effectively. In other words, AI is likely going to lead to further outsourcing not a reversal in this trend.
Who is going to build the AI tools?
If you don’t believe that all of these tools are going to be built in-house, someone still needs to build all of these AI tools and this will likely end up being the software companies! It will take time, there will be bumps along the way but just like how the winners of AI so far have been the large MAG7-style incumbents, the winners in AI software are more than likely to be the software companies that are already in the weeds at their numerous customers and are best positioned to deliver these tools and services. Again, of course there will be new, amazing companies that come up and supplant others but these will likely be the exception not the rule.
Reprogramming our online habits
This one is a bit of a tangent from our topic but we think the implications that AI will have on how we interact with the internet leaves a big question mark that will be interesting to watch unfold. Decades ago, businesses and essentially industries were created out of Google and online advertising. Digital websites could make money by creating content and selling ads on those websites. Non-digital companies could instantly grow their reach in a measurable way through digital ads such as ad words. Chat AI tools have potential to materially reshape how we interact with the internet and search. Reaching a potential customer through the ‘blue links’ at the top of a Google page might become less effective. Actually going to a website might become far less critical if Chat GPT gives us the answer instantaneously. But here is the rub, if people are not incentivized to create content on their websites (what’s the point if your not clicking ads within it for monetization), the AI tools will have a harder time staying current and training on up-to-date data. So, the two, at this stage, probably need each other to thrive. How this evolves is still very uncertain and will be an area to watch but how or if our daily interactions with ‘the internet’ evolve will have significant implications for business both digital and physical. Food for thought before AI eats that as well.
Getting back on track, there has been a lot of concerns being voiced over AI being the end of software but this type of reaction seems to be a knee-jerk one that does not take a realistic view of how businesses and the world works. In ten years, we might all be proven wrong and jobless, working for our robot overlords but until that time comes, software is probably a pretty good place to shop for some deals currently.
>>>
Forward-looking information contained in this presentation is based on the current estimates, assumptions, expectations and projections, which are believed to be reasonable as of the current date. There is no assurance that these estimates, assumptions, expectations and projections will prove to have been correct. You should not place undue reliance on forward-looking information contained herein. i2i receives no compensation of any kind from any companies that are mentioned in our analysis and commentary or on our website. Any opinions expressed are subject to change without notice. The i2i Fund, employees, writers, and other related parties may hold positions in the securities discussed in these pages, presentations, or on our website. Any information, recommendations, or statements of opinion provided here and throughout the i2i website are for general information purposes only. It is not intended to be personalized investment advice or a solicitation for the purchase or sale of securities. The information contained in this publication are obtained from, or based upon publicly available sources that we believe to be reliable. i2i makes no warranty as to their accuracy or usefulness of the information provided. i2i will not be liable for any losses or liabilities that may be occasioned as a result of the information or commentary provided. Do not buy or sell any security without conducting your own due diligence or consulting an advisor. i2i Capital Management acts as the funds exempt market dealer, portfolio manager, and investment fund manager. Introductions are permitted exclusively through registered exempt market dealers. If you are an accredited investor and interested in investing in the fund, please contact Ian Moorhouse, a registered dealing representative, at ian@i2icapital.com.
