What does generative AI mean for software companies?

The success of ChatGPT and the race among competitors to bring out their alternatives points to a growing bubble in generative AI and machine learning. The expansion of machine learning will undoubtedly change how we live and do business—in fact, it already has. The technology already helps deliver a flawless customer experience, automates routine tasks, increases productivity, and more.

But what will it mean for software companies? How will the programming industry reshape to adapt? That’s the question some of us are afraid to ask, as the race for dominance in machine learning and AI will surely have far-reaching consequences for software development.

The potential oligopoly, changes to the day-to-day running and long-term trajectory of software companies, and other aspects indicate that the recent progress in machine learning will translate into restrictions for some, but great opportunities for others.

Structural changes wrought by AI

Some of the initial changes will be felt in day-to-day maneuvers and staffing structures. AI will mean that routine and basic code writing will be displaced and generated automatically. Consequently, manual programming and the roles of coders and developers will move to creative and more sophisticated tasks.

There has been much speculation that AI will undermine the role of developers or, indeed, reduce the number of developer jobs. But it seems more likely that it will instead change the role of developers. Once AI takes over repetitive or basic tasks, developers will be able to work on the creative aspect of software, giving them more time to improve their company’s software product. AI will be able to search for gaps between technologies, which will create a need for even more software. In this way, AI could actually create more work and opportunities for software developers.

AI will also start operating in more departments of a company. Currently, many machine learning models are focused on generating code. But as technology advances, they’ll expand into more sophisticated, higher-level activities. This will include things such as product management or running marketing campaigns. AI is already creating marketing campaigns via Facebook Ads, where it can experiment with creatives and texts and get direct feedback. AI will be able to support marketers by gathering customer data, meaning that they will be able to run more effective and personalized campaigns for their customers.

Data worldwide is expected to grow to 175 zettabytes by 2025, which means that machine learning will play a crucial role in analyzing data effectively for marketers to deploy in campaigns. Much in the way that AI has taken over basic coding work, it is now able to do some of the basic marketing steps, such as analyzing or placing ads.

These are broad shifts that are likely to be seen within many individual companies. But some of the other changes that AI will bring to software developers depend on how the AI and machine learning market and oligopoly develop. This is currently evolving, and we haven’t seen the full picture yet.

The evolving AI oligopoly

For machine learning models that generate images, the current major players are the startup competitor from MidJourney and the open-source Stable Diffusion. Text is a much larger landscape. There are smaller versions of GPT-2, while everything else is proprietary. OpenAI (Microsoft) and Google have emerged as key players in this space.

It seems likely that we’ll ultimately see an oligopoly similar to the one that occurred with mobile platforms, such as iOS and Android, or with cloud providers, namely Amazon Web Services, Microsoft Azure, and Google Cloud. This will result in a few key players imposing a tax on the industry. They will be the sole infrastructure providers that everyone needs to use because the functionality they will enable will become table stakes. Consequently, rates for usage will likely become high.

OpenAI and its competitors will become key infrastructure providers. This will affect which business models may be feasible for small-scale startups and will define which types of businesses may be possible. This is similar to the way Apple’s App Store rules have made some businesses possible and others impossible. Recall that Apple rule changes removed sexual wellness, swimwear apps, certain books, and major names such as Tumblr from the App Store. 

There is a risk, therefore, that some smaller or more niche companies operating in certain markets will be cut off from using the technology or will be at a disadvantage. But it isn’t just small businesses that may be affected by this growing oligopoly. Larger companies may become strategic competitors to those advancing in the oligopoly and may become cut off from the infrastructure or unable to secure lasting advantages, reflecting previous conflicts such as those involving Amazon and AWS and their competitors.

New frontiers and opportunities

While the expansion of AI may cause restrictions, it will also create niche and specialist opportunities for startups and those looking to offer specialist services. Regionalization and potential regional laws may mean that companies operating in certain countries may need to use homegrown or local providers. Certain industries may also grow to have specific requirements for AI providers.

For example, HIPAA compliance has repercussions for the use of OpenAI, ChatGPT, or Google’s API to process patient data. Companies able to cater to the particular needs, security concerns, and regulatory requirements of particular industries will find new opportunities. The situation may be comparable to the one we’re seeing between SaaS and on-premises software and the impact this is having, such as the German government’s decision to move its primary healthcare organizations to Matrix for communications to control the key infrastructure.

AI products will give software developers more autonomy over their products and success. Products will have significant positive feedback loops, comparable to social networks. Products such as AWS cloud services improve with more users, but this is only indirectly. Amazon is able to invest more money in its infrastructure, particularly in the form of fixed costs spent on developing software products. But products such as ChatGPT benefit from the data that software companies provide it directly, which makes it difficult for competitors to catch up.

Where does this leave us?

In the future, programming will become a more interactive and exciting process. Instead of just writing lines of code all day, programmers will focus on giving instructions to computers and designing systems that work flawlessly. It’s not just about avoiding errors, but also about making sure everything fits into the bigger picture of a massive project with millions of lines of code.

But here’s the thing: AI, often misunderstood as a threat to human programmers, will actually empower them. It will take care of the repetitive tasks, giving programmers the freedom to engage in high-level thinking. It will be a boost to their creativity and problem-solving abilities. 

The future of programming is all about collaborating with computers and designing systems effectively. Programmers will step into the role of system architects, ensuring error-free and seamless solutions for extensive codebases. With the assistance of AI, they will have the opportunity to explore their intellectual potential and focus on the exciting aspects of programming. 

It’s an evolution that promises to unlock new heights of human cognition. It also promises to usher in a new era of technology industry winners and losers. Fasten your seat belts!

Kirill Skrygan is IntelliJ department lead at JetBrains.

—

Generative AI Insights provides a venue for technology leaders—including vendors and other outside contributors—to explore and discuss the challenges and opportunities of generative artificial intelligence. The selection is wide-ranging, from technology deep dives to case studies to expert opinion, but also subjective, based on our judgment of which topics and treatments will best serve InfoWorld’s technically sophisticated audience. InfoWorld does not accept marketing collateral for publication and reserves the right to edit all contributed content. Contact doug_dineley@foundryco.com.