Real returns from Artificial Intelligence

An understanding of a business' existing competitive advantage prior to its application of machine learning will aid investors in separating the wheat from the chaff - determining who wins and who loses. Our message is simple: the most predictable outcome of AI is that strong businesses can become competitively stronger, pending execution.

We are amidst a global infrastructure build out to support the coming decade of AI development. At present, hyperscale cloud service providers are investing in physical infrastructure (data centres, GPUs and networking) as they compete to bring trained AI models to market, ahead of anticipated corporate demand.

Today, most corporate adoption is being directed toward off-the-shelf usage of products (e.g., Co-pilots and SAAS applications with embedded AI), whereas internal development of proprietary machine learning models to create industry specific solutions remains in its infancy. There are countless roadblocks for the typical enterprise to work through - legacy database structure limitations; data governance, cyber & privacy concerns; budget approvals and access to specialty engineering talent - roadblocks that won't be solved overnight.

For early movers, however, the potential economic unlock will be meaningful. AI algorithms (reading & structuring copious amounts of corporate data) combined with workflow automation tools (think excel macros on steroids) has the potential to displace substantial human effort in every corporate's business processes - analogous to how the invention of the computer mainframe replaced rooms full of accountants completing journal entries in the 1980s. But as the masses of corporations catch up on AI capability, it is likely that first-mover economic unlock is merely competed away.

On the other hand, companies with existing competitive advantage have an opportunity to hold on to productivity-led margin gains, or cut product pricing for clients to widen the value proposition vs. competitors. Put another way, when a business is less subject to competitive forces, the pressure to compete away efficiency gains is less pronounced.

Where AI execution becomes more compelling, is when a leading business can leverage its existing scale, modern compute infrastructure and access to data to create new product solutions for clients that lethargic incumbents are unlikely to have the agility to replicate, and new entrants (with no data) simply cannot build.

Two recent ASX-listed Investor Days serve as interesting case studies to make this point.

HUB24

Hub24 (ASX: HUB), a leading wealth platform used by Australian financial advisors to serve their end-clients, recently showcased its plans to reduce inefficiency in a typical financial advisors' workflow.

Buried under reams of administration (e.g., both client and regulatory documentation and reporting), advisors have less time to perform the value-added functions - meeting new & existing clients - than should be possible. Using its proprietary machine learning tools to structure previously unstructured data (e.g., extracting data from PDF files, emails and audio recordings), Hub24 is building automation solutions to streamline advisor workflows.

In one example, it soon expects to be able to extract data from meeting audio recordings and pre-populate the previously laborious documentation an advisor would have to complete following a client meeting (e.g., records of advice, client file notes, follow up communications and summarised action points).

Such innovations could prove to be transformative solutions for clients. When this scope of R&D is compared to that of the incumbent wealth platforms, it’s tangible to believe a stronger competitive advantage is being built at Hub24.

Altium

Altium (ASX: ALU), a dominant electrical design software in the printed-circuit-board (PCB) vertical, recently showcased work underway to bring its vast supply chain data - underpinned by its ownership of Octopart - more directly into the component selection workflow of a PCB design. This innovation will help businesses improve electrical design planning and manage component availability risk, and by virtue of this unique data overlaid with machine learning, a non-replicable solution is emerging as part of its flagship software.

In each case, years of machine learning R&D, unique data assets, and a track record of consistent innovation create a clear pathway for investors to see how AI can (and should) enhance the value propositions of each company. Importantly, when compared to slow-moving incumbents or sub-scale peers, one can appreciate how companies such as these have an opportunity to widen their differentiation - an opportunity AI is accelerating.

The winners of the AI Revolution

At ECP, we invest in exceptional businesses underpinned by Sustainable Competitive Advantages. For us, AI has become a clear moat-extension opportunity for those that execute well, allowing strong businesses to become competitively stronger.