Never offer feedback on a demo tape. You probably didn’t come to this article today to get advice from a recording studio engineer, but here we are.
First, though, we need to intro today’s column.
- The Nick Cave Trap (436 words)
- Boston Dynamics is out of control! (144 words)
- A quantum rant (262 words)
The Nick Cave Trap
A fan recently wrote in to music superstar Nick Cave’s personal newsletter to share a song written by ChatGPT in the style of… you guessed it… Nick Cave.
The artist’s response was friendly, though it contained nothing but vitriol for the AI system.
Here’s a snippet:
“Mark, thanks for the song, but with all the love and respect in the world, this song is bullshit, a grotesque mockery of what it is to be human, and, well, I don’t much like it — although, hang on! rereading it, there is a line in there that speaks to me — ‘I’ve got the fire of hell in my eyes’ — says the song ‘in the style of Nick Cave’, and that’s kind of true. I have got the fire of hell in my eyes – and it’s ChatGPT.”
Fair play Nick, but for your fans reading this at home, here’s a little advice from a human who knows a thing or two about artificial intelligence: never critique the bot’s work.
Nick’s response nails the zeitgeist surrounding artist uproar related to generative AI. It’s hard to be an artist. Whether you’re just getting started or you’re a household name, no musician wants to hear that — in a world where harmonies, melodies, and percussion can all be digitized, synthesized and bought via a subscription service cheaper than Netflix — now it feels like lyricists and lead singers are on the brink of being automated too.
Still. You don’t critique the bot’s work. For the exact same reason that you don’t give negative feedback on a demo tape unless you plan on being involved in the artist’s future: you’re either motivating your competition or training your replacement. There’s no upside.
OpenAI is in the middle of working out a potential deal with Microsoft that would give it a company evaluation worth more than the entire global recorded music industry.
You, me, and Nick can all assume these models are just going to get better. You can rage against the machine all you want today, but tomorrow’s hit-makers will be the people who integrate powerful new tools into their art.
The electric guitar revolutionized American blues. The drum machine democratized pop music. And personal recording studio software suites such as Pro Tools created the entire independent hip hop and indie rock industries.
Not to mention, these machines are far more human than you think. They’re trained on our data. It probably makes more sense to help steer them than it does to outright dismiss them.
To quote Riley Goodside, “for now, LLMs are unavoidably products of our design, and to work they need our feedback.”
Boston Dynamics is out of control!
Haaaaaaave you seen the new Boston Dynamics video?
It’s time for Atlas to pick up a new set of skills and get hands on. pic.twitter.com/osOWiiBlSh
— Boston Dynamics (@BostonDynamics) January 18, 2023
It’s flat-out amazing what they’re able to accomplish at that lab. That’s an 89kg robot doing Olympic twist-flips (I can’t be bothered to look up the proper gymnastics term, standing by for a reader correction).
That company is out of control in the best possible way. I can’t wait to see their Cirque Du Soleil-esque show in Vegas once the company finally works out a fiery routine.
I still think that we’re a lot more than a decade away from consumer AI-powered robots, for far too many reasons than I can get into in this column today, but it’s cool to see the choreography team at BD doing their thing.
A quantum rant
I started reporting on the quantum computing industry back in 2017 because it became crystal clear to me, at the time, that the future of artificial intelligence technology would be built with hybrid quantum/classical computing architecture.
From where I’m standing, it’s difficult to imagine a pathway to artificial general intelligence that doesn’t pass through a quantum computer. I don’t think we experience base reality as a construct that can be expressed through binary logic, and thus, our world isn’t as classically computable as computer scientists would have us believe.
The reason I’m ranting about this today is, here we are half a decade later and, IMO, the quantum computing industry has made more progress than all the other STEM fields combined. Over the past five years, we’ve seen time crystals, quantum energy teleportation, and myriad other breakthroughs that would make Einstein’s hair stand on his head.
And, yet, every single quantum computing story you read is little more than a regurgitation of what I was writing back in 2017: quantum computers are expensive, you need massive cooling systems, qubits are noisy.
Newsflash: most of those problems have been sufficiently solved to the point that today’s quantum computers are genuinely useful.
I can promise you that Google, Microsoft, IBM, and the hundreds of startups and academic labs operating quantum systems today aren’t overwhelmed by the infrastructure and noise problems like they were circa 2015.
It seems like the tech journalism industry is going to get caught entirely off guard as 2023 plays out. I’m predicting massive upheaval in quantum computing by December.
@mrgreene1977 on Twitter.
Read more More Than Meets the AI here