GE 34: AI - A gift or a curse for software vendors?

Does the ubiquity and accessibility of AI tools make it easier for companies to build their own solutions and avoid investing in third-party SaaS products?

There is a choice of AI tools available now for every stage of the software lifecycle including user discovery, prototype design and code generation. Generative AI tools have the potential to totally replace certain types of software such as knowledge management, training and customer support tools.

"Generative AI can create new content and ideas, from text to images to music, and it's revolutionizing fields like customer service and training." - Gartner, Inc.

Companies looking for collaboration, workflow, document management, research, customer relationship and data reporting solutions may feel that the additional AI assistance tips the balance back onto the build side of the familiar build/buy argument. Internal solutions could potentially be built more quickly for less money and have advanced capabilities that were not possible just 12 months ago.

Such companies own vast amounts of data and documentation that could be put to use for training purposes. They have direct access to internal subject matter expertise and they could build unique solutions that may provide a competitive advantage.

"Data is the new oil. Those who own the data, own the future." - Clive Humby, Mathematician and Data Scientist

Code generation abilities allow teams to generate code for different platforms even if the internal development team is not expert in the required languages. Junior and mid-level development positions could eventually be replaced or supplemented with AI code generation.

Debugging, testing and deployment activities can also be accelerated using AI tools which significantly reduces the traditional risks associated with the build option and again reduces cost and time.

So how worried should software vendors be? Well despite all of the above, there are several good reasons why vendors have a reason to be optimistic about their collective futures.

New Technology Adoption

Established software companies are used to adopting new technology. In the last ten years, we have seen rapid adoption of cloud services, the harnessing of big data and rapid implementation on a variety of mobile platforms. Each new technology brings risks and opportunities and software vendors have the people, processes and experience to exploit these opportunities while managing the risks.

"The adoption of new technologies always brings a mix of risks and opportunities. Companies that manage to balance innovation with risk management will thrive." - Harvard Business Review

The business model of these companies relies on their ability to do this in a repeatable, predictable way and their reputations are built on how they execute against their technology strategy. The stakes are high which drives a good balance of cautious innovation.

This becomes increasingly important when solutions are large, complex and mission-critical. It is one thing to use AI for say an experimental service assistant in a single department, quite another to embed AI-driven decision making in a core part of the business that affects the entire organization.

The Power of the Crowd

Software vendors have multiple customers who can be diverse in size, location, risk tolerance and industry sector. When introducing a new technology such as AI, it is incredibly useful to survey this customer base and understand how the new technology is being introduced, what benefit is being derived and what lessons have been learnt. The term best-practice is perhaps over-used in product marketing circles - but when you have hundreds of customers who have an AI project in play there is a lot of information available.

Software vendors are able to learn quickly. What is working? What is not working? Where are the real risks and where is the real value? When this learning is shared, it can save an individual company weeks and months with their own implementation. A facility that is just not available when building a solution in isolation.

The Safety of the Crowd

For highly regulated industries such as Financial Services and Healthcare, there is also a level of safety in being part of a software vendor community. Sure, bad things can and do still happen, but they happen to a collective. The industry benchmark falls temporarily, but issues are addressed quickly as the vendor and the community pool their resources to respond and rectify issues. From a regulatory point of view, an individual firm is not isolated and is less likely to be heavily disciplined if the issue is prevalent across the industry.

The +1 Effect

Artificial Intelligence is powerful of course, but by itself is of limited utility. AI applications need a collection of reliable data from which to work and need to be plugged in the right places in the end-to-end workflow. The right users need to be provided access and those users need to be trained and provided support. This “application infrastructure” is not trivial to build and frankly is not as interesting to build for internal development teams as more in-vogue technology such as artificial intelligence and machine learning.

"AI enhances existing capabilities, making processes more efficient and data-driven decisions more accurate." - Accenture

However, mature software platforms already have these features in abundance. More importantly, practically every one of a platform’s existing capabilities can be augmented with AI. This is the +1 effect. Data ingestion processes become more flexible and powerful. Documentation can be automatically generated and easier to search. Integrations become easier to implement and test. Reporting and analytical tools are able to answer more sophisticated business questions. Forecasting and predictive tools become viable, easier to deliver and more accurate.

Software vendors can take all of their existing expertise, domain knowledge and platform capability and combine this with advances in AI to deliver these +1s. And it is the nature of AI implementation, that many of the enhanced capabilities take less time to develop than traditional new features.

Conclusion

AI can be a gift for software vendors provided they understand their role in its future story. Vendors can add the most value by applying their years of experience and software engineering discipline. They can ensure that AI is being implemented in a safe, ethical and reliable way that is fit for an Enterprise’s mission-critical applications.