My best 2017-2018 articles

Most social network posts are meant to be read and forgotten in a week. But some longer articles keep as a useful read even after years. Here are my best articles in 2017 and 2018:

I devoted the first half of 2017 mostly to Artificial Intelligence. This meant studying hundreds of pages of mostly linear algebra (matrices) with Python language machine learning libraries. If you wondered: the Python's name is derived from the British comedy group Monty Python, programmers can be odd. Unfortunately, the AI reality above did not match most of the AI marketing (this microwave oven uses AI, this refrigerator uses AI, we will replace all the workers with AI, etc.) and some journalists let people be afraid of the Terminator robots coming soon. So I wrote 3 very easy to read articles for a wide audience, trying to avoid any of the hype and inaccuracies so common in marketing and press media:

1) AI (Deep Learning) explained simply.

2) Will AI kill us all after taking our jobs?

3) Deep Learning is not the AI future.

The articles were published also in newsletters for AI researchers and got good feedback from scientists and practitioners too, the 3rd article won the "most shared of the year" badge on KDnuggets (over more than 1000 posts) and was translated in Chinese. Proudly, in this case, I could write articles appreciated by both the AI guru and the average curious person. In 2018 I did not write more about AI, because there were no real major AI discoveries.

Next, we got the Bitcoin and crypto hype. Really, another one! I was trading Bitcoin back in 2011 when we were a handful and I saw a few crypto bubbles already, but every a few years new people enter the game to make the same errors. In 2017 when the Bitcoin price was rising and everyone was launching a blockchain startup with token crowdfunding (ICO), I warned how most would fail (like the Altcoins years before) and how letting the average person invest was a suicide: [ICOs must add VC-like selection, else it's a crowd-suicide]. The article is long and comprehensive, if you read it all you will know the same or more than the millions of self-claimed "blockchain experts". I helped some honest and good faith blockchain ICO startup founders, but unofficially: I didn't allow to associate my name to any ICO, I knew most was getting legal or reputation issues. In the article [ICOs can be fixed with automation, not with guidelines for humans] I gave my advice to those who tried to self-regulate, without success since the space was fragmented and flooded by scammers and lunatics too quickly. In 2018 my predictions were confirmed, the crash in value and reputation of the crypto stuff was as crazy as the 2017 hype, so I had nothing to add. Until Nouriel Roubini's testimony for the US Senate, where he said the blockchain is just a scam and not even a tech thing, just a normal database. That's incorrect, so I wrote a couple of small articles as a response: [My comments on Roubini's "Crypto is the Mother of All Scams and (Now Busted) Bubbles"] and [Why the blockchain is useful but not for the applications you hear about]. In short, people and the press either said the blockchain is the future of everything (nonsense), or that it's a complete scam and no tech (really?), the truth is something in the middle, but balanced reviews are boring and not published or shared.