AI is often portrayed as a binary innovation: a force that will divide man and machine in untold and sometimes frightening ways.

But as the panellists at our recent event, "The future of Enterprise Software: The age of AI" discussed, AI may well represent an opportunity for founders and enterprises to come together and use this ever-advancing technology to unlock unprecedented advantage.

Here are my key takeaways from the discussion between HSBC’s Group CTO, Mario Shamtani, and Tom Hulme, Managing Partner at Google Ventures.

The possible impact of AI on large enterprises is best understood by focusing on different user sets.

The vision for users

Users will benefit from the opportunity to co-create systems that replace redundant, repetitive work presented on responsive, intuitive interfaces – primarily powered by voice input.

The vision for departments

Large Language Models, LLMs, will only get more sophisticated and specialised – with the potential to reshape entire departments. Legal, accounting and even HR – teams with large volumes of repeatable, manual work, will be able to focus on exceptions to the rule rather than dedicating all their time to processing every query.

The vision for customers

Call centres and help desks that are designed to help address common problems will become far more effective, with operators only required on rare instances – leading to vastly improved, personalised customer experience (CX)

So, what could that mean for organisations more broadly?

There are 2 clear areas of impact:

1. The ability to rapidly prototype and test solutions will empower organisations to meet growing customer expectation.
2. Teams will have more time to focus their expertise on solving complex problems (aided by ever-improving tools).

Bottom line, business leaders will have new tools to deliver superior experiences at a fraction of the cost, reducing complexity while maximising efficiency and productivity.

Although change is upon every enterprise, there are certain principles of modern enterprise infrastructure that may remain true.

Data quality remains paramount

While every IT-enabled endpoint is a potential source of data that can help core LLMs improve over time, the size of the opportunity is directly proportional to the quality of data being fed into the system.

Diligence and compliance are crucial

Organisations carrying technology debt may find it difficult to embrace the opportunities on offer because of legacy data issues. Similarly, new platforms will also need to be carefully managed and maintained, particularly in regulated industries – arguably more stringently than ever given the fact that there is little to no precedent for AI to manage governance without supervision.

It’s clear that it’s the large organisations that are well structured to build on their current foundations that are poised to benefit – but how best to do so?

The opportunity that AI-led innovation brings is clear but the race between founders building solutions, and enterprises leveraging them, is a bit like the tale of the tortoise and the hare.

On one side, there are founders building solutions at break-neck speed off LLMs that are becoming exponentially better the more data they ingest.

They are undoubtedly moving far more quickly than their ideal customers – large, established organisations. Unlike agile startups, most large enterprises have neither the speed nor the ability to incorporate innovation at this level (particularly in highly-regulated environments).

It’s not a lack of desire to adopt new technologies; it’s just that the reality of implementing change at this kind of scale is complex. Without careful change management, implementing AI too quickly could translate to significant technical and cultural debt. Think progressive pilot programmes, not overnight change.

For Mario, this disconnect means that it will be crucial for the enterprise and founders to form long-lasting partnerships and co-create relevant solutions.

Introducing people with a deep understanding of their organisation and its needs to visionaries who are fluent in front-line innovation is a powerful way to find solutions that are scalable and safe without the cost of hiring scarce, top-tier talent.

It’s as close to a win-win as you can get. Enterprises have access to solutions that can give them the edge, while founders can get the insider view and adapt their products accordingly, closing the gap between perceived and actual product needs.

Unsure about finding the right balance? A simple litmus test is not only to think about shared vision, but to question whether the relationship is one that will see neither party cannibalised, or both boosted by mutual benefit.

There’s no crystal ball that will allow us to accurately predict the full impact that AI may have on the enterprise space, but the perspectives from each of the key players is telling.

The VC view

"Right now we’re in a rational bubble", said Tom. Despite outward appearance, he suggests that picking winners in the space remains an art, predicting that many investors will likely lose despite the fact that "anybody with an IT endpoint falls into the TAM" and everything from coding to Legal will be impacted by AI-led innovation as part of the "distributed revolution".

As he so eloquently said, it’s about “unlocking the undigitised economy”; and the founders that identify high-volume, high-impact opportunities to ease enterprise suffering are well placed to succeed.

He also suggests it’s companies that can attract talent and choose not to throttle their efforts given perceived cost that will be at the forefront, with their position constantly strengthened by improving models and buyer comfort levels.

The enterprise perspective

Marco envisions a world where zero-cost platform management means there will only be a need for event-driven responses, freeing up time for improved customer interaction and a focus on strategic imperatives.

Interestingly, he did stress that there are certain systems – core infrastructure – that may well go largely unaffected due to the cost and complexity of building new foundations.

For founders in the enterprise space, the goal is clear: stay at the forefront of AI as it develops, and then find a way to translate those new possibilities into meaningful solutions.

Not every enterprise will need generative AI to solve issues – but they could all likely benefit from enhanced efficiency brought on by replacing manual, repetitive tasks.

Build something fundamental that is difficult to copy but easy to implement, that really moves the needle for big companies, and you will be well poised to succeed.

Ultimately, it will be the innovators who can navigate the complexity of large enterprises and go beyond ideas to show measurable impact that will capture the imagination of the biggest fish in the pond. Once those large enterprises mobilise, the world will truly change.