DEVOPSdigest

Generative AI is revolutionizing software development, with 92% of developers in a GitHub survey saying they use AI in their work, and 70% saying they see advantages to it, according to Chris Wysopal, Co-Founder and Chief Security Evangelist at Veracode.

"I think it's close to universal now, and I don't see that changing," adds Mike Loukides, VP of Emerging Tech Content at O'Reilly Media.

DEVOPSdigest invited experts across the industry — consultants, analysts and vendors — to comment on how AI can support the software development life cycle (SDLC). In Part 8 of this series, the experts express opposing views on the current state of AI adoption across the development community.

AI Trending Right Now

"I feel that AI adoption is already on a steep exponential curve," says Ed Charbeneau, Developer Advocate, Principal, at Progress. "GitHub claims that over 20k organizations have already adopted Copilot. The promise of productivity increases due to AI tools guarantees that the trend will continue."

There's a lot of excitement around AI right now that will only push adoption further, observes David DeSanto, Chief Product Officer, GitLab. "In fact, according to our 2024 Global DevSecOps report, 78% of respondents said they are currently using or plan to use AI in software development in the next two years, up from 64% last year. The use of AI for software development is quickly becoming table stakes, even among those who were initially hesitant."

In addition, Sterling Chin, Senior Developer Advocate at Postman cites the 2024 Developer Survey by Stack Overflow, stating "76% of respondents said they are using or are planning to use AI tools in their development process." And "81% agree increasing productivity is the biggest benefit that developers identify for AI tools. Speeding up learning is seen as a bigger benefit to developers learning to code (71%) compared to professional developers (61%)."

"Whether we like it or not, the AI revolution is upon us, and so over the next few years we'll continue to see a growing rate of adoption across the board. I venture to guess that almost every aspect of an engineer's workflow will be impacted by various AI tools," Chin predicts.

We expect to see the widespread use of AI across the board in innovative companies, including AI-powered code completion and suggestion tools, the emergence of AI-powered debugging tools, and a greater use of AI for project management and planning, adds Dotan Nahum, Head of Developer-First Security at Check Point Software Technologies.

Increasing Experimentation

"For many organizations, we believe that 2024 is about laying the foundation for AI adoption," says Patrick Doran, CTO at Synchronoss.

Igor Kirilenko, Chief Product Officer at Parasoft, believes more and more software engineers will adopt AI tools in the development process, and they might start their journey with simple experiments to see what AI can do or just to learn how it works, but as they better understand the value those tools can provide to increase developers' productivity, the adoption of AI will only increase in the next years.

"We're still on the on-ramp and feeling our way," Jon Collins, Analyst at Gigaom adds. "AI for augmentation is an absolute boon, understanding where to do better. However the next couple of years should be seen as a phase of active experimentation."

Slower Than We Think?

Despite the stats cited above, however, some experts hold to the opposing viewpoint that AI adoption is not growing as fast as some believe.

Jeremy Burton, CEO of Observe, says, "These trends always take longer to adopt than we in tech ever imagine. GPT is certainly a phenomenon and passed 100M users quicker than any other technology. That said, I'd estimate that AI tools in software engineering have less than 10% penetration at this point."

"Based on what we've seen in the first half of 2024, the adoption of AI tools for development remains pretty low," Michael Webster, Principal Software Engineer at CircleCI, concurs. "This isn't surprising, considering how other disruptive technologies have entered the market before becoming mainstream. New technology just takes time to adopt, especially for developers who are very comfortable with their existing tools and stack. More seasoned developers are especially prone to skepticism, while junior developers are open to more experimentation. That being said, while I expect adoption will remain flat for 2024, in the next few years, we'll most likely see AI become more integrated into software development workflows as a few generations of new grads get familiar with the tools."

Arguably the usage of AI in software development is still niche or experimental, Chris Du Toit, Head of Developer Relations at Gravitee, asserts. While some developers are no doubt leveraging it's capabilities, few organizations have developed a corporate-wide development methodology with formal practices on the use of AI.

Roadblocks to Adoption

The belief that AI is not taking hold in the development space as rapidly as suspected could be based on some of the roadblocks presented by the experts:

AI Limitations and Risks

Those who have not jumped on the AI bandwagon cite concerns such as: generative AI code lacks the creativity and innovation of human developers, AI can produce code with security vulnerabilities, and code developed with AI may contain errors, bugs, or inefficiencies, according to the Reveal 2024 Top Software Development Challenges survey from Infragistics.
Casey Ciniello
App Builder, Reveal and Slingshot Senior Product Manager, Infragistics

Finding the Right Tools

One challenge is adopting the right AI and AI tools. Similar to the mid 2000s when everything was iThis or iThat, not everything that is "AI" is good. Being able to analyze the productivity gain for engineering teams is a hurdle that many engineering leaders are not prepared to analyze. Leadership, whether at the top, or at the team level, need to fully understand where AI tools can fit into the workflow of the engineering teams. If the tool doesn't reduce the menial and the mundane work of an engineer, it won't be adopted. This is where the "fad" aspect of the industry will struggle. "AI is just a fad" will be shouted from the rooftops because tools that were purchased simply didn't solve the core problems for the engineering team and were never adopted. But the right AI tool will be lauded as a hero.
Sterling Chin
Senior Developer Advocate, Postman

Implementing AI into Existing Workflows

Many teams are struggling to integrate AI into their existing workflows, leading to inefficiencies and suboptimal outcomes. Automation and orchestration play crucial roles within the larger generative AI ecosystem. The lack of skilled professionals who can properly implement and manage AI-powered workflow solutions further complicates the situation. As a result, businesses are often unable to fully leverage the potential of AI to drive innovation and achieve success.
Eoin Hinchy
CEO and Co-Founder, Tines

The Need for ROI

The innovations in the code AI space in the past 18-24 months have proven one thing yet again — new technologies sometimes come with a lot of hype and see fast adoption but smart enterprises always go back to demanding true ROI on their investments. The dominant companies in the code AI space are already witnessing the trough of disillusionment from their customers who adopted code AI early. Now, the conversation is less about "What if I am left out of this AI race?" but more about "Does this AI truly unlock opportunities for me better than what we can do today?" Many products that were early movers or became internet sensations are now struggling to prove this ROI to their users/customers. On the other hand, companies that prioritized enterprise customer value from the beginning are now benefiting from the increased sophistication of code AI evaluation by customers.
Raman Sharma
CMO, Sourcegraph

I expect significant investment, particularly in software intelligence, where AI can understand the code base, metrics, and developer productivity. However, it will be challenging to prove the ROI of AI tooling. It all sounds great, but it's difficult to measure how much more productive a developer or a team is without slowing them down.
Rupert Colbourne
CTO, Orbus Software

AI Adoption Expected to Accelerate

New technology is always met with some resistance — developers may be hesitant at first, but with continued education to help them understand how it will make their jobs better (vs. replace them), organizations should be able to increase the usage of AI and automation for tangible results, says Matt Healy, Director of Product Marketing, Intelligent Automation, at Pega.

"I would characterize AI adoption today as healthy, but still very early," notes Jeff Hollan, Head of Applications and Developer Platform at Snowflake. "Teams are starting to enable it and are seeing the benefits, especially when it comes to writing code. In time, I think it will accelerate drastically to the point where development efforts at the end of the decade might even take place without some AI assistance in the loop."

"I expect that nearly all developers will be using AI for code generation (where permitted) in the next few years," Peter White, SVP of Emerging Products, Automation Anywhere, forecasts. "Not all tasks are amenable to AI but, for those that are, the benefits of AI are impossible to ignore and allow organizations to significantly speed up their development processes."

Adoption is expected to grow rapidly as organizations face intense pressure to accelerate business transformation to optimize productivity and output in an evolving economic environment while also contending with a shortage of skilled developers and the ever-growing IT backlog, confirms Jithin Bhasker, GM & VP for the App Engine business at ServiceNow. For example, Gartner predicts that "by 2028, 75% of enterprise software engineers will use AI coding assistants, up from less than 10% in early 2023."

With this in mind, the upcoming developer ranks may be the ones to truly embrace AI.

Neha Goswami, Director of Engineering for Amazon Q Developer at AWS agrees, "I see the new generation of developers, who are coming out of college, are already very used to looking at these generative AI-powered software development assistants to assist their research tasks or learn new languages. They have already made a shift in the way they are doing development."

AI Advancements Drive Adoption

With AI-assisted development still in a relatively early stage, many experts expect to see even more amazing advancements that will drive adoption further.

In 2024 and beyond, AI tools are expected to become integral to software development, according to Rahul Pradhan, VP of Product and Strategy at Couchbase. As these solutions become more prolific, major IDE, application lifecycle management vendors and cloud providers will integrate and further expand their AI offerings, making it easier for developers to leverage AI with tools and coding languages they're already familiar with. This will drive an uptick in AI-assisted code generation, maintenance, security testing and more.

"As AI advances in the next few years, automation will be capable of taking on more duties across the development process, and able to fulfill more specialized and specific use cases. The developer experience of working alongside AI will also improve and become more collaborative and seamless, further spurring adoption," says Shomron Jacob, Head of Applied Machine Learning & Platform at Iterate.ai. "I expect AI tools to become increasingly integrated into the software development lifecycle. Similar to how version control systems are now ubiquitous within developer toolsets, I see AI coding co-pilots becoming just as mainstream."

The trend of new emerging developer-focused AI tooling will surely continue, Matej Bukovinski, CTO of Nutrient agrees. "With so many companies focused on this technology, I see no other outcome. There are still more advancements possible even without major machine-learning breakthroughs by just further refining LLMs, working on efficiency and building additional integrations into various developer tooling. With increased reliability, lower costs and better integrations into everyday processes, it will become increasingly easier for companies to justify investing in this kind of tooling."

AI - Must-Have

Advances in AI capabilities and the need for competitive advantage will further accelerate adoption, making AI a strategic necessity for companies looking to optimize costs and speed up time-to-market, says Pradhan from Couchbase.

AI coding assistants will become essential parts of software development every day, removing the burden of mundane, repetitive tasks, confirms Harry Wang, VP Strategic Partnerships & Head of Product Marketing at Sonar.

Engineering teams need to understand that AI is here to stay. It's not going anywhere, Chin from Postman proclaims. Adoption will continue to increase over the next few years. As in the early 2000s, if a company didn't have a website, it wouldn't survive. Today, if teams don't believe AI is useful, they'll quickly be surpassed by teams that do adopt AI.

Go to: Exploring the Power of AI in Software Development - Part 9: Workforce