Date: 2025-10-05 Page is: DBtxt003.php txt00025847
AI ... ARTIFICIAL INTELLIGENCE
STANFORD HOFFMAN-YEE GRANT SYMPOSIUM
The Hoffman-Yee Grant Symposium highlighted the work of the 2022 winners of the Hoffman-Yee research grants
STANFORD HOFFMAN-YEE GRANT SYMPOSIUM
The Hoffman-Yee Grant Symposium highlighted the work of the 2022 winners of the Hoffman-Yee research grants
Original article: https://www.youtube.com/watch?v=IzK55L26FgA&t=7835s
Peter Burgess COMMENTARY
This is a 5 hour Youtube video of a Stanford AI symposium. I have not watched it end to end, but have dropped in and out for almost the same amount of time. It is fascinating, but at the same time seriously ... or dangerously ... underwhelming.
I have been interesting in 'how the world works' since I was quite young ... perhaps about 7 years old when I found a collection of bound magazines that appeared after my parents closed down my grandfather's house. There were about 10 years worth of leather bound Strand Magazines and about the same number of Harmsworth Popular Science volumes. They were from the 1890s and 1900s prior to WWI. I spent hours and hours thumbing through these magazines absorbing all sorts of information that was in the news 50 years before I was born.
This may explain why during my adult life I have always been interested in the context of whatever I am observing, and especially what I am so crtical of so much of the 'instant news' that is served up thanks to amazing tecnology in a continuous stream every day of our modern lives.
Part of my formal training was doing 'articles' to qualify as a Chartered Accountant in London in the UK. My articles were from 1962-1965 with Brian Maynard, a partner of Cooper Brothers & Co who headed up their London consulting practice. At the time this training was far and away the best training for business management available in the UK which had not yet embraced the 'business school' cult that was getting popularized in the United States by the likes of the Harvard Business School and the University of Pennsylvania Wharton School.
Though in many ways the modern Chartered Accountant is dangerously 'out-of-touch' with much of the modern world, the central role of accountancy as a key component of modern corporate management has not degraded significantly, though in my opinion, it is on 'borrowed time' and essentially on 'life support'.
It is noteworthy that in 5 hours of this Stanford HAI Hoffman-Yea symposium on AI, there is zero reference to accountancy. From my point of reference, this is surprising ... and shocking ... because it suggests that the accountancy profession is out of touch with reality and progress as well as Stanford also being in something of an 'academic bubble' somewhat removed from the real world.
Some ot this reminds me of my own early career. When I was in the twenties, I got a job working for a company called Aerosol Technique (ATI) in Milford, Connecticut. My title was Treasurer and I was essentially assistant to the CFO. The CFO had graduated from the Harvard Business School after professional training as a Chartered Accountant in London. Interestingly, even though he had a Harvard Graduate School MBA, he had not passed through any undergraduate university program either in the UK or the USA!. My first assignment was to 'fix' the company's budget system that had been launched the year before and had proved a complete disaster ... a program headed up by another Harvard Business School MBA with zero practical experience. Though I was young, my experience working and training at the same time to qualify as a Chartered Accountant in the UK proved to be a very effective way of learning real world practical management as well as the underlying core of the subject. Though the system was manual and 'rather clunky' it was orders of magnitude better than anything that had gone before.
Soon after this I was handed another corporate crisis to fix. In 1965, the company had installed an IBM 1401 mainframe computer in its Eastern Division also in the Milford, Connecticut complex. The compueter's main task was to manage a very fast moving inventory ... all the raw materials, the production schedules, the finished goods inventory. The system ... in theory ... also kept track of the location of the inventory in the warehouse. Bottom line, however, is that the computer was completely useless ... the system simply did not work ... at all.
My 'position' in the Eastern Division was 'Eastern Division Controller' responsible for all the accounting functions including running ... that is fixing ... the computer system. Nobody in the division liked the computer ... trusted the computer ... and to keep the division running, there were all sorts of 'work-arounds'. All sorts of 'consultants' from IBM and from Harvard had engaged with the problem, and there was layer upon layer of 'system' that added complexity but nothing that improved anything.
Bottom line ... the computer worked ... the system didn't.
I had been introdued to 'management by walking around' in the UK at a company called Davy United in Sheffield where I had been part of a 'Management Training' program. There were about 20 of us all of whom had good engineering degrees but little or no practical industrial experience. We were immersed in every aspect of factory work ... no books ... just learning from people who had decades of actually doing the work. I did not stay with this company even though it had some very good ideas and practices, there was something missing that was not being fixed ... MORE ON THIS ANOTHER TIME.
At ATI, I walked around and around the warehouse and the factory to try to understand the way things worked in practice. My conclusion was that everything that mattered in the warehouse moved a half dozen times in the course of a shift ... both the parts warehouse and the finished goods warehouse ... while the computer paper-work and the processing was on a once-a-day update. This time disconnect meant that it was impossible ever to get the factory and the computer to be in the same time zone.
To demonstrate that the computer actually worked and could process business information I arranged for the computer to handle the division's fixed asset accounting records. While the inventory was changing several times a day, fixed asset records changed very very slowly ... maybe a few times a year! In short order, we were able to impress everuone with the computer listing of all the fixed assets of the division with details of cost, peiod depreciation, accumulated depreciation and net cost.
Next we had to figure out how to do something useful related to the material flows and the production cycle. It was pretty obvious that the locator system was not at all fit for purpose and never would be with a paper based movement ticket system and daily updates ... but the core system could be 'tweeked' in order to deliver reliable information from the warehouse and production lines to serve the needs of the accounting and sales admin departments. Fro asles admin, daily information that was accurate was a najor step forward and for the accounting department getting everything right once a month was good progress and essentially good enough.
While I was doing all I could to get a handle on the paperwork flow around the computer system, a young materials manager / warehouse supervisor was doing his own thing to sort out the warehouse problems. He was both a solution and a problem because he wanted a warehouse that looked efficient because this is a good first step to it actually being efficient. But the system paperwork could not handle by any stretch of the imagination, the movement that he was making happen. Essentially ... any semblance of meaningful paperwork disappeared together with all the information needed to administer production planning, requirements planning, and so on. This had to be resolved urgently to keep the factory working ... and to his credit 'he got it' and quite soon a practical solution was found to ensure the office to what it needed accurately and on time.
Much of this saga at Aerosol Techniques was documented via the Harvard based Management Analysis Center (MAC) consulting firm to become a series of Harvard Business School cases related to business computerization.
I started by observing that the AI experts seem to be ignoring the role of accountants in the corporate world. And I should add that I am equally concerned that accountants seem to be ignoring and misunderstanding the role of AI as they talk about the future of accounting. My impression is that accountants are ignoring how accountancy could be applied in a meaningful way to all the key progress and performance metrics in the socio-enviro-economic system. Rather, it seems that accountants remain locked into the idea that everything that matters revolves around investors and the growth of corporate profir.
The TrueValueMetrics.org framing of progress and performance metrics is simply that social and environmental issues should have the same weight as economic (profit) issues and it is the blend of socio-enviro-economic progress and performance that should be getting better and better at the maximum possible rate.
Peter Burgess
Stanford HAI
14.5K subscribers ... 644 views
Sep 20, 2023
The Hoffman-Yee Grant Symposium highlighted the work of the 2022 winners of the Hoffman-Yee Research Grants. The teams presented their results to date, plans for the future, and competed for additional funding of up to $2 million over the next two years.
The grant program is a multiyear initiative to invest in research that leverages artificial intelligence to address significant scientific and/or societal challenges aligned with Stanford HAI’s key areas of focus: understanding the human and societal impact of AI, augmenting human capabilities, and developing AI technologies inspired by human intelligence. We believe the results of these projects could play a significant role in defining future work in AI from academia to industry, government, healthcare and civil society.
At the Symposium, the 2021 recipients of Hoffman-Yee Research Grants also presented results from their research to date and plans for the future.
Transcript