Industry executives and experts share their predictions for 2023.  Read them in this 15th annual VMblog.com series exclusive.

Eye on AI: Why 2023 is a turning point for mass adoption

By Anmol Bhasin, CTO, ServiceTitan

We're all familiar with some of the buzziest AI innovations of the last decade: from autonomous driving, facial recognition, gene sequencing and first generation AI chatbots to new entrants such as generative AI routines (i.e. ChatGPT, Dall-E, Stable Diffusion, and Vall-E). 2023 will continue the innovation streak and we will see an acceleration of developments in the generative AI space that will fundamentally alter the creative workforce as they will likely see their roles transform from artistic creators to astute curators.

In 2023, three themes will stand out. First, innovation will continue its breakneck pace. Generative models will etch away and large technology corporations will enter a battlefield of smarter and transformational modules along chatbots, digital imaging, voice assistants, intelligent workflow automation agents, diagnostic imaging in healthcare, and more.

Second, this pace of innovation will force tough conversations in board rooms and executive meetings. Last year, strategic conversations at companies centered around digital transformation, and taking a bite of the AI apple. In 2023, the conversation will shift to "why are we behind?" or "have you tried ChatGPT?". Instead, executives will start asking, "what will this do to our industry, our workforce, and our customer's business?" After the initial phase of bewilderment, astute leaders will search for opportunities to leapfrog, and essentially bypass a whole generation of AI applications, entering the brave new world of third wave or "generative" AI.

Finally, the enthusiasm at the executive level of organizations paired with macro-economic softening will motivate companies to identify and implement opportunities for AI and automation. This paradigm will become the new mode of operation to drive margin efficiencies, and increase resilience amid economic uncertainty as well as future proofing themselves. Concerns around workforce replacement and the fears of AI "taking over" will become much more subdued in developed economies but even more so in developing ones.

An environment for mass adoption

Technological advancements often exhibit a virtuous cycle - innovation leads to refinement, which improves adoption, and ultimately leads to increased demand for new innovations. With AI, we've refined the way we use the technology for a variety of use cases, making it much easier for non-experts to make the case for adoption. The technology to fully automate a lot of business processes exists and continues to improve, while new AI templates and plug-and-play solutions are creating the perfect environment for widespread adoption in 2023.

There's a construction analogy I like to use when comparing the AI of the past to that of the present: in the past, we built AI brick by brick using toolkits and packaged Software Development Kits (SDKs) to build models. Think lots of cement and mortar in the form of data. It was time intensive and we needed experts to ensure every brick was laid with precision. Now, instead of bricks, the industry has evolved to use prefabricated homes that you can place and deploy in a day. Cloud software vendors continue to invest in AI modules and value added services that have reduced the friction to build and adopt solutions using AI for highly repetitive and labor intensive tasks.

Let's dive into the current dynamics pushing businesses toward AI adoption:

1. Economic conditions and a strong talent pool. With the current economic climate, particularly inflation and a competitive talent market, we'll see an explosion of AI and other technologies to automate roles that companies will struggle to fill while watching their bottom line. Automation can cut unnecessary spending to improve a company's runway and improve efficiencies to work faster and stay competitive. Many businesses today already have useful data that could be more easily analyzed with AI to allow for a streamlined workflow, comprehensive risk analysis, customer support, and workforce optimization - they just need to have a set of application strategists and integration practitioners to integrate their systems with the offerings provided by numerous vendors.
2. Configurability of AI modules and influx of use case templates. A majority of cloud providers are offering easy-to-operate AI templates allowing businesses to simply add new tools to their existing technology suite. First, the configurability of these modules is becoming mainstream, especially as it pertains to customer support. For instance, large language models can be trained and constrained to provide answers from a limited knowledge base pertinent to a company's product line. If deployed correctly, the majority of the support interaction can shift over to chatbots powered by large language models. Second, SaaS companies are now providing out-of-the-box, customizable hyper-specific use case templates. For example, an AI recommendation template, now available for the trades industry, automatically recommends what parts should be on a service truck as technicians headed out to the field every morning. In the next five years, there will be thousands of templatized workflows with embedded AI and automation capability. We could also see more businesses using some form of AI to automate workflows and increase efficiencies.
3. Decreased barrier to entry. While AI experts and data scientists have been in high demand in recent years, we're going to see many more AI "promoters," "operators," or "managers" crop up this year and beyond. This is a significant and recent change signaling that mass adoption is underway. With mainstream, free-to-use, generative AI services like ChatGPT for text creation or DALL-E for AI imaging, we're also seeding a new generation of potential AI natives. As these AI templates continue to gain popularity and the barrier to entry continues to lower, we should expect to see widespread adoption across industries and an emerging workforce that is more experienced with AI systems.
   

Where to expect AI adoption

So what areas of business are ready for "prefabricated" AI capabilities? We'll see this transformation take place primarily where human bandwidth can be replaced or enhanced, and where we've seen successful or easily-scalable use cases.

1. The front office. We've seen that the first to adopt automation are people in the "front office" - the roles or departments with a direct tie to the customer or end-user, such as customer support, marketing, and sales operations. Advancements in AI language models have improved chatbots, while AI-powered data analytics can deliver powerful customer insights.
2. Some middle office jobs. IT, compliance, and risk management teams are also ripe for automation - we've seen standout use cases from the banking and finance industry showing promise for others in the space. Businesses will look to automate tasks that involve interpreting dense, disparate data, so human workers can focus on analyzing or elevating insights for both clients and executives.
3. Specialized industries with existing use cases. Organizations are also more likely to adopt AI with existing use cases, more easily predicting the business return on investment. This makes sense, as most organizations will use tools with proven success over something new and buzzy. With the proliferation of templates and new use cases, we'll see many specialized sectors or roles using automation where the ROI is significant.
   

It's worth mentioning there are still industries where AI will be more difficult to implement and experts will still be needed to be heavily involved (i.e. brick by brick, not prefab). These industries are much less likely to be disrupted by current AI technologies as implementation is more costly and there are currently fewer use cases. Complex back office systems, such as finance, accounting, and operations, will still struggle to apply practical AI solutions and the back office will take much longer to build and adopt working AI programs. Similarly, human-centric jobs, such as HR departments, are unlikely to see large-scale AI adoption in the very near future.

What does this all mean?

Of course, the question on many people's minds is: Will AI replace me or my job? Legitimate concerns over job security and the fear that AI will replace humans has previously prevented AI from advancing more quickly. However, I believe these fears are misplaced, and before long, more of us will be able to see the major opportunities that AI present.

As Deloitte pointed out in their research on the workforce and automation, "ue to automation, over 800,000 jobs have been lost but nearly 3.5 million new ones have been created." The question is not how AI will replace us, but instead, how will we collaborate with AI to work smarter, faster, and push the boundaries of human performance? Many of us will be answering this question sooner than we think.

##

ABOUT THE AUTHOR

From Linkedin to Groupon to Salesforce, and now as CTO of ServiceTitan (the cloud-based software platform for the trades), Anmol Bhasin has a long history of building transformational cloud-based products. Prior to his role at ServiceTitan, Anmol worked on a battery of LinkedIn’s data products including Job Recommendations, Ad Targeting, LinkedIn Talent Match candidate targeting tools. He then led Salesforce’s venerable Einstein products and platform weaving in AI across the company's sales, service, and marketing business lines. Most recently, Anmol spearheaded the launch of Titan Intelligence — software that brings the power of data and AI to the trades on a scale that has never been seen before. In addition to spearheading the launch of Titan Intelligence, Anmol is responsible for leading global Product and Engineering and pioneering new SaaS technology to improve the lives and businesses of contractors across the country.