AI Investment Landscape: The Potential of Smaller-scale Success

In our previous article, we outlined ways in which Onfolio can leverage AI to grow the company and build value for shareholders.

It’s well worth going back and reading the article, but in short, we plan on::

  • Adding AI to existing businesses
  • Buying businesses and adding AI into them for immediate improvements
  • Using AI to create free customer acquisition tools
  • Acquiring or building standalone AI projects

It’s important for us to point out that we see AI as an augmentation of our existing strategy (which you can read here), rather than us suggesting we are now an AI company.

AI is powerful and can do many things, but ultimately we believe AI will eventually just become a built-in feature of software and a tool for creating better software in the first place…rather than a standalone product. People will use Google Docs and they’ll work better than ever because they’re AI-powered. People won’t say “I want to use a new AI document software”.

What we find to be the most exciting about AI – the ability for us to work it into our existing businesses to improve them – might seem less exciting than some of the other hype around AI.

But we actually think when it comes to creating unicorns and VC-investable businesses (which isn’t our strategy anyway), AI may turn out to be an anti-climax.

Many current AI-based products primarily operate by leveraging API calls to existing AI platforms such as OpenAI’s GPT-4, and then applying a thin layer of what’s called “prompt engineering”.

Despite its widespread use, this approach results in a weak technological moat that might be more ephemeral than we think. As AI evolves, it learns from these prompts, a phenomenon that can be likened to a “self-draining moat”. This is described well by Tyler Tringas in this tweet thread.

As Tyler mentions, the value derived from these AI models will most likely either directly benefit the platforms that power them or become dispersed across boundless competition, which will diminish potential profit margins.

We often see a new tool released that can do something incredible, but then we think “That will become a feature of Zoom, or Google Docs, or Microsoft Office and people won’t know or care if it’s AI.”

In the long run, consumer value is not determined by whether a tool uses AI, per sé, but by its functionality, utility, and ability to deliver the best value.

Thus, the shiny object syndrome of AI might eventually give way to a more pragmatic evaluation of utility and performance.

As mentioned above, in the context of AI integration, major incumbents like Adobe, Microsoft, Google, Notion, and Zoom will likely dominate their respective AI-enabled markets.

The primary barrier for these behemoths is the sheer size of their user base and the costs associated with implementing AI on such a large scale.

This scenario brings into focus the potential challenges for the traditional VC model in the AI sector. With high implementation costs and lower margins, the opportunities for absorbing large amounts of capital and generating multi-billion-dollar value may be limited.

This is a stark contrast to the conventional VC model that thrives on deploying large capital for potentially huge returns.

Despite these challenges, the AI landscape is far from barren. There is still some opportunity for those willing to chase smaller (but by no means small) wins.

A more modest exit of say, $50m, can still be considered a remarkable victory in the grand scheme of things, especially if the process to build the business was not too intensive and it ends in a strategic acquisition by larger firms looking to bolster their AI capabilities. Some of the behemoths mentioned above will likely end up innovating via acquisition, as is often the way.

Moreover, as more professionals harness AI to increase productivity—coding being a prime example—there are increasing opportunities for small teams to challenge large incumbents using limited capital. This is an area where Onfolio has been experimenting and developing tools of our own.

The market is ripe for entrepreneurs who can take a small team and a little capital to build fully-fledged competitors to some of the more frustrating-to-use incumbents.

This shift is a potential setback for large VC funds as these companies require less capital, have leaner business models, and aren’t necessarily inclined toward the traditional VC fundraising approach.

What does this mean for Onfolio’s approach?

Well, let’s review what we’ve stated so far:

As a collection of entrepreneurs used to bootstrapping, being nimble, and focusing on small-scale operations, AI has created some interesting opportunities for us. Software development costs and speed have come down massively, and created opportunities for us to develop AI tools quickly, and cheaply.

While we are still honing our skills in the space, we are already generating new ideas for products or services.

For example:

Our Proofreadanywhere audience stated that using the Chicago Manual of Style can be a pain, and is not easy to reference. At the same time, tools like ChatGPT do an admirable job of proofreading, but often lack the nuance or judgement that a human proofreader needs to complete a higher standard job.

This gives us an opportunity to a.) Create a proofreading tool that makes it easy to reference CMOS and other style guides, and b.) Fine-tune a proofreading tool that is much better than the 80-20 “good enough” quality of general-purpose LLMs.

Or another example – StableDiffusion and Midjourney create excellent graphics, but can they create logos, graphics, fonts, or icons to the same level that we sell on MightyDeals.com? With access to data (such as the 15+ years of sales data we have), one could build a more specialist tool tailored to graphic designers, that would allow a greater quality and wider scope.

Finally, in the field of Search Engine Optimization (SEO) – one we are very familiar with, you can ask basic questions to ChatGPT about how to rank higher in Google, but only a specialized tool specifically trained on the nuances and finer details of SEO will be able to recommend winning strategies.

None of these tools are likely to become unicorns, but we don’t need them to be.

Can they become lucrative tools (and perhaps acquisition targets for the incumbents) in their own right, and can we create them without an expensive development budget and large team?

Absolutely.

This is why we are giving serious thought to creating a dedicated subsidiary (ex: Onfolio AI LLC) and completing an equity financing round for said subsidiary, raising a few hundred thousand dollars, up to a million, and using those funds to rapidly complete the various tools we have planned, plus more that come to mind. We are sure if we do this we can create value for both our existing subsidiaries, spin off new tools, and ultimately generate value for Onfolio shareholders.

If you’re interested in potentially making an equity investment into the Onfolio AI subsidiary, fill out the form below and we’ll follow up with you should the time come.

Onfolio AI SPV interest form here

AI is creating a lot of opportunities, but where the value flows is still unknown. We believe it is less likely to flow towards the creation of new billion-dollar companies, and more towards feature roll-outs from the existing giants. Below that though, there is still likely millions of dollars of value to be captured by those who are satisfied with smaller, cashflow-positive businesses – exactly the realm we operate in.

Sign-Up to Receive Investor Updates Such as Filings and Press Releases