Person
2 episodes · 40 quoted claims
Tom Davenport said a survey he ran with MIT CDO IQ and AWS found that about 80 (apparently percent) expected generative AI to transform their organizations, but only six had a production application.
“80 said they thought it was going to transform their organizations, but only six had a production application that they had deployed.”
Tom Davenport said DBS, the biggest bank in Southeast Asia, had about 35 generative AI use cases but none in production, largely for regulatory reasons.
“And they said, Yeah, we've got like 35 use cases, but none of them are in production yet, largely for regulatory reasons.”
Tom Davenport said Randy Bean's annual survey of data leaders had long shown figures in the low 20s on having a data-driven culture, down from the 30s in some cases a few years earlier.
“Have kind of bounced around in the low 20 area for years, and it's even getting worse. In some cases, it was in the 30s a few years ago.”
Tom Davenport said Randy Bean's latest survey, after twelve years of surveys, more than doubled the share of data leaders reporting a data-driven culture.
“in the latest survey that I think will be published by the time this podcast is out. It more than doubled these figures to organizations.”
Tom Davenport said Marriott staff told him about a Tripadvisor review saying the pool was too cool, which older sentiment analysis could not interpret as sarcasm.
“They said somebody wrote on Tripadvisor the pool was too cool, and you know, how do you interpret that?”
Tom Davenport said the advanced data analysis feature of ChatGPT works by writing Python code to do the analysis.
“The advanced data analysis part of ChatGPT works by writing Python code to do all of that stuff, so it's really quite astounding.”
Tom Davenport said a two-line prompt to a generative AI system can produce about three pages of machine learning model creation, analysis and feature engineering.
“a two line prompt will get you three pages of machine learning model creation and analysis and feature engineering”
Tom Davenport said OpenAI had a little over 300 employees, far fewer than Google's many thousands, yet still beat Google to market with generative AI.
“OpenAI was sort of an interesting example because they had 300 and something people employed there, compared to what 80 zero at Google, or maybe it's even bigger, but they still beat them to market.”
Tom Davenport said OpenAI is now closely tied to Microsoft.
“Generally, of course, now they're in bed with Microsoft in a big way, so that I was going helps them.”
Tom Davenport said BMW is training a very large number of white-collar workers (transcribed as '80 zero', likely 80,000) in citizen development.
“BMW training 80 zero white collar workers on how to do citizen development.”
Tom Davenport said his next book is about the citizen movement, in which non-professionals take part in system development, automation, analytics and AI.
“my next book is about the citizen movement, you know, non professionals in system development and automation and analytics and AI.”
Tom Davenport said he is writing his next book with Ian Barkin, who comes from an automation background.
“I'm writing this book with Ian Barkin, who he comes more out of an automation background.”
Tom Davenport said chief information officers are encouraging citizen development because the queues at IT departments for building applications are getting longer.
“the lines at IT for developing applications are getting longer, not shorter.”
Tom Davenport said he co-authored an MIT Sloan Management Review piece arguing that AI is a poor field for fast followers because accumulating data takes a long time.
“I co-authored a piece with Vikram Mahidar in SMR a number of years ago about AI. Is not a good area to be a fast follower in, because it takes a long time to accumulate the data that you need”
Tom Davenport said a Stanford study found little cheating with generative AI in schools.
“It was a study out of Stanford saying that there's not much cheating happening because of generative AI in schools.”
Tom Davenport said some Babson College faculty asked the academic technology office whether AI could be shut off on campus.
“some actually went to the academic technology office and said, Can we shut off AI on our campus?”
Tom Davenport described a copywriting experiment with three conditions: a plain word processor, an advanced type-ahead system with accept-or-reject suggestions, and full generative AI output.
“One was just a word processor, no generative AI, the 2nd was a sort of a advanced type ahead system where you could accept or reject”
Tom Davenport said that in the experiment, copywriters preferred the intermediate type-ahead tool, which produced higher quality work and more online click-throughs.
“And in general, the people, the copywriters, liked the intermediate stage more, and it produced higher quality work and that got more click throughs for online.”
A Home Depot employee used Alteryx to produce sales forecasts in a few hours instead of weeks, according to Tom Davenport.
“he was able to produce a sales forecast in a few hours at most, rather than the weeks that it took him previously.”
Home Depot cited extra income well into the nine figures from the forecasting automation, according to Tom Davenport.
“They cited, um, uh, well into the nine figures of extra income as a result.”
The Home Depot forecaster was named employee of the year and later became a consultant helping others use such tools.
“he got employee of the year. Now, as often happens, he's no longer doing that. He's now a consultant to help other people use these tools.”
A BMW employee told Tom Davenport the company is training 80,000 people in citizen development approaches.
“we're training 80,000 people in citizen development approaches, so it's really, um, going viral.”
Low-code and no-code tools began emerging roughly five years before this episode, letting non-programmers build small systems, according to Davenport.
“a couple, maybe five years ago, you started to see these quote low-code, no-code tools emerging that would let you build small systems”
Robotic process automation vendors introduced easy-to-use tools that let people build their own workflow automations.
“the robotic process automation vendors introduced tools that were very easy to use, and people could create their own kind of workflow automations.”
A former head of data science at a large European bank told Davenport there was opposition to citizen development there, which Davenport thinks was reasonable.
“he said, you know, there was some opposition, and, you know, there probably should be.”
PwC has a group of mostly technically amateur employees called digital accelerators who are charged with taking the firm into the digital age.
“And they have a a group of people who were amateurs for the most part, technically, who, um, have been charged with taking them into the digital age and digital accelerators, they call them.”
PwC encourages its digital accelerators to submit their applications to a central hub where they are evaluated.
“And they're encouraged to submit all of the applications they develop to some central, uh, hub, and in that situation, they're evaluated.”
PwC pays contributors a small amount depending on how much their approved application is used.
“And they even, you know, pay a little bit to the to the person who contributes it, depending on, you know, how much usage it gets.”
At Lego, the IT department was initially unsupportive of a demand forecaster's automation, but the head of demand planning supported it.
“Initially, the IT department was not supportive, but the head of demand, uh, planning was supportive of it.”
Davenport conducted a second annual survey with the MIT Chief Data Officer Group Symposium, sponsored by Amazon Web Services.
“I was just looking at a survey that I did with the, um, MIT Chief Data Officer Group Symposium, the second annual survey sponsored by Amazon Web Services and data just came in.”
Some surveyed companies reported that no use of generative AI was authorized, which Davenport considers a bad idea.
“there was no authorized use of generative AI in their companies. I think in general, that's a bad idea”
In the survey Davenport cited, only eight percent of organizations said they had generative AI systems in production.
“only eight of organizations said they had systems like this in production.”
Davenport says companies are making deals with providers like OpenAI or Google so their prompts do not enter the providers' models, addressing intellectual property concerns.
“We don't want our prompts to make it into your model. That solves a lot of the problems relative to intellectual property ownership”
Code Interpreter was a beta offering from OpenAI at the time that could perform machine learning analysis from a simple prompt.
“there's a generative AI version of that called Code Interpreter. Um, it's a beta offering from OpenAI now”
Davenport reports that a two-line prompt to Code Interpreter produced three pages of data analysis using a machine learning model.
“two-line prompt gave me three pages of data analysis with a machine learning model applied to it.”
Johnson & Johnson applies criteria to volunteers for its citizen development program, such as data experience and technical aptitude.
“J and J has a s a series of criteria that they apply to you once you volunteer.”
Deloitte offers certifications for citizen developers and citizen data scientists, with levels of training.
“they have certification of citizen developer, citizen data scientist, and so on.”
Davenport recalled that AT&T said it had saved 13 million minutes through its automations.
“I think it was ATT that said, Oh, we've saved 13 million minutes from our automations.”
A PepsiCo finance leader sponsoring citizen automation told Davenport the company tries to determine the economic value of each automation, such as costs avoided or hires prevented.
“every one we try to figure out, what's the economic value of it? Is it costs avoided? Somehow? Is it, you know, a hire that was prevented?”
Davenport expects broad-scale citizen development adoption at BMW once its 80,000 people have been trained.
“I think it's gonna be happening at BMW when they've finished training their 80,000 people in it.”