As organizations continue to learn about the opportunities in machine learning and AI, they look for business cases where to implement these new capabilities. As typically happens with new tech cycles, companies can directly access the lower levels of the stack using these capabilities. As time goes by, vendors and software solutions that solve common problems appear.
Software companies are developing bundles of Gen AI solutions (Microsoft with Copilot, Adobe with Firefly, Salesforce with Einstein GPT, ServiceNow with Now Assist, etc.) to solve lower-level issues that enterprise users need to be abstracted from.
One of the relevant concerns from corporate decision-makers around large language model adoption (LLMs) (also called foundation models) is the hallucination limitation, which means that generative AI platforms create confident responses that cannot be grounded in any of their training data. Leaders cannot tolerate implementing a solution that may spit out incorrect, erroneous, or made-up information for their teams to use (not even considering any potential bias that may have been included in the training data).
With this consideration, we may see most software companies developing highly performing AI features (some that may be marketed as AI and others not marketed as AI), where they solve user pain points more seamlessly. As it happened with the internet before or with smartphones afterwards, a software company that does not deliver features via the web (obvious today), or does not have a mobile app where users can interact with the service (of course!), can become irrelevant in short order.
As we previously wrote mid-year in https://alten.capital/blog/generative-ai-impact-on-tech-services, we believe the new machine learning and generative AI capabilities will transform productivity, especially in high GDP per capita countries.
At Alten Capital we invest in the technology services sector. Please contact us to explore a partnership.