AI in a weekend

For all the criticisms the current wave of AI receives for being a one-trick pony, this article neatly summaries the counterpoints. There’re incredible AI application opportunities available with the current state of deep-learning, as illustrated in this article. We are just getting started.

https://www.forbes.com/sites/robtoews/2020/02/09/deep-learning-has-limits-but-its-commercial-impact-has-just-begun/

Deep Learning Has Limits. But Its Commercial Impact Has Just Begun. Deep learning may be bumping up against conceptual limits as a model of intelligence, but opportunities to apply it to transform industries and enact massive real-world change still abound.

"AI can spot cancer better than humans…” We see these headlines, but is that necessarily a good thing? Especially when we don’t clearly know what cancer really is? Given a bad cell, it could just sit there and do nothing, could spread slowly causing no real harm, or could spread rapidly and kill. When we say “detect cancer” don’t we mean detecting the kind that could kill? If that’s the case, we currently don’t have a way to know how the bad cell will turn out.

One could argue isn’t it better to spot bad cells early so that we can monitor it? There’s a real challenge with that - Over Treatment. Once we hear we may have cancer, how many of us would not take preventive actions, actions that could potentially cause more harm than good. This article below brilliantly outlines this issue.

The over all point is, just because "AI can spot cancer better than humans” doesn’t really mean that it’s going to help us beat cancer better, at least until we make more breakthroughs in the science of understanding rogue cells better.

https://www.theverge.com/2020/1/27/21080253/ai-cancer-diagnosis-dangers-mammography-google-paper-accuracy

Why cancer-spotting AI needs to be handled with care Finding more cancers isn’t always the answer.

AI in a weekend added a Shop on website button to their Page.