Innovation advisory and IRM/InfoSec/I-Audit services. Positive contrarian, inspirator, loosen-upper.
No, this is not about bias in data that you train your “AI”i.e.ML on. I’ve posted (not nearly) enough about that. It’s more about pointing to an HBR article that should be(come) very influential… As it was already known that any ‘training’ was about as ineffective as one could get, qua e.g., security awareness transfer … Continue reading “Training your way out of bias”
Yes, yes, there’s still discussion about biases and how ‘Ethical AI’ should be pleonastic. 1. Though it isn’t, and inherently can’t be. For one, since ‘ethics’ is (factually) about discussing and debating dilemmas ad nauseam without providing a clear-cut decision – whereas human behaviour of the non-pathological kind is about something else… For another, any … Continue reading “Biasin’ -out.”
Word has … had it, that continuing doing what didn’t work, would be futile. (Remember?) And, there was a thing called the Recency Bias (as, somewhere, in here). Against the first, some point out its truth (quod non), that outcomes wouldn’t change. But when the environment (say, the actual world) in which the ‘experiment(s)’ were … Continue reading “Inverse Recency Bias”
This will be published in the July ISSA Journal. Just put it down here, already, to be able to link to it. ;-] And, first, a picture: [Toronno, ON] After which (Dutch version linked here): You have the Bystander Bug One of the major pluses of open source software is that anyone, even you, could … Continue reading “(Ahead of time, because we can)”
Hey, I know what you were looking for, so here it is: The full set of Bias Time lists. Just as a dump, with typos and all. First, a picture: [Going, Going Gent]
[Terug naar de hoofdpagina] Iets meer over mijn ‘vakhobbies’, behoudens wat daarover behoudens… ondergenoemde links allemaal reeds op m’n blog stond: (Kwantitatief Operational) Risk met natuurlijk een stevige -Managementcomponent; over wat Doug Hubbard (hier en hier) te berde bracht, aangevuld met deze, deze en deze bij voorbeeld. Voorbeelden te over… Denken vanuit strategisch perspectief: welke … Continue reading “Vakkig”
[Voor een eerdere Engelse versie van deze post zie hier – maar het issue is nog steeds actueel, blijkt] De meesten van u zullen het wel hebben meegekregen; het resultaat van een groot onderzoek van MIT[1] over beslissingen die een Moral Machine zou moeten nemen om aan te sluiten bij wat wij mensen moreel handelen … Continue reading “Oud, tevens nog steeds nieuw: Auto-mobielen voor non-ethici”
“Individuals can be ethical, groups not so much.” — as a conclusion of this insightful piece. At which, immediately my Canetti and Ortega y Gasset antennas went off the charts. And with the note of the ‘can’ in the above — not so much the petty little bureaucrats with their mostly mumbo-jumbo about cold fusion … Continue reading “Group ethics”
Since reinforcement learning is too broad a term… we’d better call it continued-in-deployment learning. Back to the subject, being this CiDL. And how one would audit a system that uses, as one of its many (sic) parts, such a module. The weights change constantly, a little. By which point would one say that the system … Continue reading “DevML Auditing”
In a slight pre-view of an upcoming post about how to do ‘AI’ right in any organisation [hint: not bottom up per se but rather top-down with b-u as a sauce…], there is this little gem about “Efforts fall short in the last mile, when it comes time to explain the stuff to decision makers“. … Continue reading “Walk the data then Talk”