Interesting idea from @joinmassive: instead of paying media and apps with advertising and user data, people can pay with opt-in compute cycles. Mine crypto, train #AI, distributed tasks like three SETI program, etc. via an SDK that developer can adopt. > https://t.co/lk3PQ1lbkq


$120B in spending for 2022, 20% growth from last year.

Imagine a future where #AI can generate these hypercasual games. Initially, we’d see oversaturation. Then, hypercustomization: “Hey Siri, today I want to play a game with huge animated chess pieces.”



Go @Grammarly! Another favourite in my Startups Worth Watching list: https://t.co/wS2YvG5shf

I used it for over 5y now and I couldn’t imagine working without it. >

Grammarly raises $200M to expand its #AI-powered writing suggestions platform https://t.co/YbYNx6FrOn


It’s amusing that, out of 14, 4 libraries are about #AI and 2 are about data mining. In 2017, I started suggesting that a new generation of automated and adaptive cyberattacks would come to life. This is just the beginning. https://t.co/TpqQIFVkXZ


Must-read article on the cost of training to #AI in 2021:



Occasionally, to have some alternative fun during the weekend, I’ve started programming in Python. Of course I’m focused on #AI applications. After learning some data mining via Selenium, I spent some quality time on @spacy_io, a great NLP library. Shame I can’t install it on iOS


One of my favourite shows when I was a kid was Knight Rider.
Elon said (and showed) that Tesla effectively has one of the most advanced #AI around.

Just saying… https://t.co/fwrxQRuoIF


Every year @Adobe shows more impressive steps towards a synthetic reality. Or, as I call it, a post-real era. All this technology, intentionally or not, is key to achieving realistic #VR >

Adobe’s Project Strike a Pose uses #AI to pose your models for you https://t.co/cmG1wbPGRm https://t.co/cW6ULhpMP0


@rationalpsyche Very different problem. It exists and it’s amply debated. The one I’m interested in is barely discussed. #AI as a killer of #diversity (assuming it’s a correct hypothesis, which I don’t know) is not something you read very often.


I’d be interested in exploring a problem I call “The convergence towards a single taste”. Today #AI optimizes for content that is appealing to most. Peer pressure pushes the ones that are more edgy to converge as well. The same algorithms then effectively silence non-conformance.