3 ms·
I think your summary is good. Not too off topic: since the 1980s, I have referred to symbolic AI (expert systems, non-neural NLP, etc.), neural networks, and n
by mark_l_watson 5y ago
I think your summary is good.
Not too off topic: since the 1980s, I have referred to symbolic AI (expert systems, non-neural NLP, etc.), neural networks, and now deep learning models as “AI” and use AGI when referring to progress towards “real AI”. I only mention this because so many people now use the term machine instead. After 40 years of using the term “AI” I don’t feel like changing.
re: the article: the real story is a (allegedly, from the article) cheating contractor selling snake oil. Part of the life cycle of machine learning models is the evaluation of checking that the model stays valid in time, looking for data drift in the original training and test data, etc. That these models NEVER passed verification, is repugnant to me.
- dragontamer 5y agohttps://www.theonion.com/revolutionary-new-driverless-car-requires-zero-function-1847048521 https://www.theonion.com/revolutionary-new-driverless-car-re... > In what many are calling the future of the automotive industry, a revolutionary driverless car unveiled Tuesday requires zero functional technology to generate profit. “While many companies in the autonomous vehicle sector have unveiled models to generate funding with little feasible technology, what sets X-Course apart is that we make money without any self-driving technology whatsoever and absolutely no plans to ever develop it,” said X-Course CEO Patrick Medill, adding that his company had shattered industry experts’ previous belief that companies could only succeed in the self-driving space by at least pretending to develop new technologies.
- hef19898 5y agoBeing a writer in the comedy / satire field sure got harder in the last years...