For the 111th episode of Burning Platform, we host Frank Scavo and Dave Wagner of Avasant Research. With Computer Economics, now part of Avasant, they can boast a long track record of tracking technology trends and benchmarking various aspects of IT spending, and staffing.
Here they summarize findings from a survey of early Generative AI projects. The executive summary available for download here describes it as their “first survey-based report on generative AI (Gen AI) strategy, spending, and adoption” and it provides “metrics on how much enterprises are spending on this ground-breaking technology, where they are spending it, and how they are governing it.”
In last year’s episodes of Burning Platform (see index of all episodes here) I hosted a number of guests who covered various GenAI use cases, economics, risks and other perspectives. This is much more field research based on input from nearly 200 companies across 6 industries – healthcare, retail, manufacturing, professional services, BFSI (banking, financial services and insurance) and IT Services and Solutions
They highlight several unique angles on GenAI projects
At 10.50 they discuss governance models – are these projects centralized, are they being run by IT or federated, is the governance different by industry etc.?
Starting around 17.40 they discuss use cases by industry. While many of the early projects are somewhat horizontal and focus on customer service, procurement and HR, I liked that they broadened their focus to cover revenue angles and industry operations, not just SG&A. Several use cases are fairly ambitious – including one which described a complete replacement of an internal procurement function.
One major aha – lots of companies view software development as their most exciting GenAI use case. More vendors should be talking about impact on their implementation, application management and maintenance efforts but likely are nervous that customers will want those savings to be passed along.
Around 26.15 they switch focus to risks from GenAI.
One big finding from the survey - very little concern about job losses and in fact, expectations of job increases. I interpret that as healthy – it comes from a focus on revenue use cases, from project ownership high up in enterprises and the realization that automation tends to target individual tasks, and transform jobs as against completely replacing the human element. Media, academia and politicians tend to cling to the job loss point of view. We also invoke my book, Silicon Collar, which looked at a century of automation and included many examples such as how UPC codes led to an increase in CPG and grocery jobs, how ATMs did not eliminate banking jobs right away etc.
The GenAI space has been moving very rapidly as the episodes last year showed and I hope Frank and Dave keep expanding and updating this benchmark every few months. Enterprise buyers, for a change, appear to be ahead of enterprise vendors in terms of expectations from an emerging technology. And hyperscalers and vendors like Nvidia, Google, Meta and others keep pushing the boundaries of the “art of the possible” and moving beyond AI to AGI.
They would welcome more use cases and other GenAI perspectives for the next edition. Please free to reach out to either of them on LinkedIn
Very informative 37 minutes.