Becoming AI Native – Step 1: Mindset and understanding why “AI Native”?

Goldman Sachs Research estimates that worldwide (on-balance-sheet) capex in AI for 2026 will exceed $1 trillion with The Wall Street Journal adding an estimate of further hidden (off-balance-sheet) commitments running up to $3 trillion. The planned Initial Public Offering (IPO) of OpenAI are estimated at around $1 Trillion (Forbes) and for Anthropic around $2 Trillion (Reuters) – although neither is yet underpinned by a legally compliant IPO-prospectus showing the underpinning calculations.

This reflects an enormous expectation that AI will deliver a level of benefits to end users that justifies the cost of this new infrastructure (which is already more than global IT spend). Following the “Core Macroeconomic Paradox”, a company spending around 6% of its revenue today on (AI and) IT would need to double this to 12%, requiring revenue to jump by 50-60% and margin to increase from 15 to around 20%! Where will that level of revenue increase come from?

This has raised the concern that AI is heading for a “boom and bust” cycle, with strong challenges around the macroeconomics and use of “circular funding” which exacerbates the problem.

Should we then discount AI as just another hype?

In my view certainly not! Global Finance Magazine highlighted two levels of seemingly contradictory evidence:

1. At the enterprise level, very few companies (<10%) can demonstrate positive returns for their AI investments, visible in overall company earnings.
2. A majority of companies do report significant benefits in customer service, throughput time, cost and quality including risk reduction, and will increase their spending on AI.

This dichotomy in evidencing the benefits stands in contrast to the level of investments described above!

In my view, we are still at an early stage of AI adoption and application, with new tools emerging and being proven in practice. The major investments outlined above centre around the “magnificent 7” US based players who are betting that “Large” models are the only way to success. If so, then large models require such large investments that only large players can play…

Is gradual, incremental adoption also beautiful?

The theory that only large models will succeed is not yet widely accepted and smaller models – particularly from China – have made a dent in this. Let us not forget that the widespread adoption of microchips in the Technological Revolution (from about the 60/70's) used simpler, cheaper chips e.g. in fridges, cars and toys to accelerate mass adoption.

So whilst predicting whether the large models win and an AI bust is avoided remains difficult, we can see that the adoption of AI in key areas has already proven to be very beneficial in customer service, effectiveness and efficiency.

The “AI Native” Mindset?

Given the benefits of AI that are already proven in multiple functions, processes and business activities, executives already face the imperative to look at where they can also benefit and encourage a broad and multi-faceted adoption of AI. This may not take the form of a single killer enterprise-wide application, but should embrace multiple, bottom up or focussed initiatives, with discovery of domains where a major benefit can be harvested without risking the company on unproven, new technology.

This approach has been labelled “AI First” or more broadly, “AI Native”! Encouraging teams across an organisation to look for incremental and innovative applications of AI in multiple areas across an organisation, in addition to exploring a small set of 'big bets'.

In future articles I will highlight where to look for AI benefits, how to prepare your organisation for this and how to manage this portfolio of AI initiatives and investments to achieve success.

For now the summary is: start becoming “AI Native” today in how you look at work!