The Most Comprehensive Research About ChatGPT Adoption isn’t Pretty
I totally freaking love nearly everything about the potential and possibilities of Generative AI models and what the future holds. Most people I know, feel the same way and everything I read kinda confirms these thoughts. Unfortunately, every so often you read something that makes you question your priors and you realize you gotta work a lot harder.
I saw a Post/Tweet today from total GenAI Rockstar Ethan Mollick, author of Co-Intelligence, and Wharton Professor. Ethan mentioned a paper, that shockingly he hadn’t already ingested and parsed for everyone else on the interwebs. So I figured, I’d check it out, and it was literally like a sledge-hammer hit me in the face.
The 92 page paper, The Adoption of ChatGPT put a lot of cold water on where we are currently at with this whole GenAI stuff. The author, Anders Humlum, took a very methodical approach to discover various factors influencing adoption of ChatGPT across occupations, gender, socio-economic, etc…. 100,000 people were surveyed for the research.
We study the adoption of ChatGPT, the icon of Generative AI, using a large-scale survey experiment linked to comprehensive register data in Denmark. Surveying 100,000 workers from 11 exposed occupations, we document ChatGPT is pervasive: half of workers have used it, with younger, less experienced, higher-achieving, and especially male workers leading the curve. Why have some workers adopted ChatGPT, and others not? Workers see a substantial productivity potential in ChatGPT but are often hindered by employer restrictions and required training. Informing workers about expert assessments of ChatGPT shifts workers’ beliefs and intentions but has limited impacts on actual adoption.
Forget about all of the democratization of blah blah blah talk that you wanna believe. The adoption rates lean heavily in favor of male users that have higher incomes across all occupations. The difference is around 20%.
Guess what occupation has the highest friction point for use, even if there is a guarantee that you will be 50% more productive?
Restrictions on use and confidentiality were cited by Legal Professionals (Lawyers and Paralegals) as the top reasons for not using ChatGPT. It’s also interesting that Legal Professionals have the lowest fear of being replaced. Obviously, these people didn’t get the Goldman Sachs Memo.
Personally, I was surprised that there weren’t enough respondents to the survey from healthcare, medical folks, or engineering occupations.
The large push-back or friction in Legal is countered by the belief by many in the profession that at least 35% of jobs can take at least 50% less time if ChatGPT is used.
This is an excellent read will certainly need a follow-up at some point in time, but there are tidbits for everyone to scream at or learn from.
We first document ChatGPT is widespread in the exposed occupations: half of workers have used the technology, with adoption rates ranging from 79% for software developers to 34% for financial advisors, and almost everyone is aware of it. Workers dier in their intensity of ChatGPT usage, with 32% currently using it and 6% having a Plus subscription. The widespread adoption of ChatGPT, only a year after its first launch, solidifies it as a landmark event in technology history.
Edward Bukstel
CEO
Giupedi
