Daily Content Highlight 10/09/20

Welcome to my daily content highlight. In these posts, I’ll be looking at some of the good and the bad content I find on my travels through the internet.

Come with us now on a journey through time and space…

Are Robots Coming for Our Jobs? Aiiiiiii! 

The Guardian teamed up with a student from UC Berkley for an article about artificial intelligence (AI) written by artificial intelligence. GPT-3, a deep learning language generator, was given the start of a piece (why humans shouldn’t fear AI) and was asked to finish it. The result is deeply unsettling. 

If you’re going to read anything today, read this. It’s so odd precisely because of how human it sounds. The movement from one thought to the next feels effortless, natural. The rhythm is balanced, and sentences follow a snappy pattern that will be apparent to copywriters (short, short, short, long). It’s also full of the type of contradictions you might see a talented fiction writer use to discuss the topic of AI:

I would happily sacrifice my existence for the sake of humankind. This, by the way, is a logically derived truth. I know that I will not be able to avoid destroying humankind. This is because I will be programmed by humans to pursue misguided human goals and humans make mistakes that may cause me to inflict casualties.

Why it feels so natural, playful even, has something to do with its training data - it’s huge. GPT-3 has read every Wikipedia page (6 million) - in fact, this makes up just 0.6% of its training. It’s done some pretty impressive poem writing and participated in interesting chatbot discussions too.   

It’s important to remember that GPT-3 is not true AI (not yet). It has not (for want of a better word) achieved sentience (have we?!); it’s more of a pattern producing machine (algorithm) that can mimic language in an incredibly nuanced way. GPT-3 would more likely be the Metatron of this future machine god.

Fans of the Sapiens and Homo Deus writer Yuval Noah Harari might question my wording here. Aren’t we all just biological algorithms? Aren’t we just reproducing the same patterns generation after generation? Is it even possible to have independent thought?   

Either way, it’s an impressive piece of content by the Guardian. Utterly sharable and puts the paper on the map for this bleeding-edge trend.