Vern Glaser

Is using commercially generated big data a way to build new theory? Is refusing to use it a sign of scholarly integrity? Lindebaum et al. & Glaser et al. offer differing views. Intro: bit.ly/4awu3Yi Point: bit.ly/3JiboDG Counterpoint: bit.ly/3vRKNtP

The dark side of AI’s promised efficiencies: artificial intelligence can use data and algorithms in a way that prioritises rationality over values such as fairness and quality of education, @UAlberta’s @glaserv writes on #THECampus bit.ly/3DtBpg3

The hype around ChapGPT masks an infrastructure that depends on humans to tag training data — for pennies — and a tech stack that is so costly to buy and operate that the scale required for break even is hard to imagine. theverge.com/features/23764…

@oomidvar @mitsmr 5/If you're interested in the academic research that inspired this article, take a look at our @JMS_Journal article, which is open source and freely available. onlinelibrary.wiley.com/doi/full/10.11…

1/Omid Omidvar (@oomidvar), Mehdi Safavi and I have published a new article in @mitsmr about how organizations using algorithms and artificial intelligence can fail to effectively navigate risk and disruptions. sloanreview.mit.edu/article/predic…

@oomidvar @mitsmr 4/We suggest that organizations can combat algorithmic inertia through four corresponding practices: exposing data and assumptions; periodically redesigning algorithmic routines; assuming that the model will break; and building bridges between data scientists and domain experts

@oomidvar @mitsmr 3/We identify four sources of algorithmic inertia: buried assumptions, superficial remodeling, simulation of the unknown future, and specialized compartmentalization

@oomidvar @mitsmr 2/Some highlights: Building on an in-depth historical case studies of how Moody’s used algorithms to rate mortgage-backed securities before the 2008 financial crisis, we seek to understand what we call algorithmic inertia

"Are we ready for #AI to raise the dead? ... This is not remotely the first time we’ve faced a we-can-but-should-we moment as a society ... It may even spawn a new concept: the right to remain dead." esquire.com/news-politics/…

I'm looking forward to discussing our paper, The Biography of an Algorithm (with @lucianadadderio and @neilpollock), with the Practice and Process reading group in a couple of weeks (on 5/26)! Registration here: eventbrite.com/e/practice-and…

For those of you doing research on technology entrepreneurship, the West Coast Research Symposium hosted by the University of Washington is a great venue to present research...here's the call for papers! thewcrs.wordpress.com/call-for-paper…

This is an interesting look at the dark side of #AI from #UAlberta researcher @glaserv

Nice new paper by Dean Shepherd, Stella Seyb, and @profgerrygeorge on how instantiating business models as formal, cognitive/linguistic, and real/physical boundary objects can help entrepreneurs identify and overcome business model incoherence. (1/3) journals.aom.org/doi/full/10.54…

A great article on @SakuMantere that's worth reading! forbes.com/sites/karlmoor…

Important piece by @glaserv and colleagues @tineadam @mehreenashraf91“The potential devastating effect of using algorithms: rationality as preferred conduct and efficiency as preferred end taking over all other values, becoming ultimate values in and of themselves”

@tineadam @mehreenashraf91 @mitsmr 6/If you're interested in the academic paper published in AMR that inspired this paper, it's available here (open access) doi.org/10.5465/amr.20…

1/Dirk Lindebaum, Christine Moser @tineadam, Mehreen Ashraf @mehreenashraf91 and I have just published a new article in @mitsmr of relevance to people and organizations interested in how #ArtificialInteligence solutions get deployed in organizations. sloanreview.mit.edu/article/when-a…

@tineadam @mehreenashraf91 @mitsmr 5/To prevent this erosion of values, organizations can follow 3 principles: (1) Beware of proxies and scaling effects, (2) Strategically insert human interventions into your algorithmic decision-making, and (3)Creating evaluative systems that account for multiple values

@tineadam @mehreenashraf91 @mitsmr 4/The potentially devastating effect of using algorithms: rationality as preferred conduct and efficiency as preferred end taking over all other values, becoming ultimate values in and of themselves

@tineadam @mehreenashraf91 @mitsmr 3/This induces a process that we call the mechanization of values, in which everything is forced into a quantifiable straightjacket and we limit consideration of the full spectrum of human values

@tineadam @mehreenashraf91 @mitsmr 2/Some highlights: - When we integrate algorithms into business processes, we tacitly agree that everything must be measured, quantified, standardized, and rationalized so that it is ready for computation