Why TRUST is essential for AI to be successful [Chris Castle]
With UMG leading the major label charge to purge most AI-generated music from streaming platforms, Chris Castle explains how the rise of AI could affect all musicians and why building a trusting relationship with the technology is vital.
by CHRIS CASTLE of Music Tech Policy
We are told that artificial intelligence is a powerful tool that may end up being either the end of humanity through automated super soldiers making autonomous decisions regarding their own AI devised rules of engagement, or life saving medical procedures and diagnostic tools like House meets HAL. As usual–both outcomes are probably equally likely if humanity doesn’t keep the deus in the machina. We really don’t want them thinking “Hell is other machines.”
The question I have is how will we keep humanity around when companies like Google are hell-bent on achieving the Singularity ASAP. This is particularly true of creators–let’s not kid ourselves that the Google Books project was some altruistic motivation to build the digital library of Alexandria rather than a massive digitization project to build a large language model to train artificial intelligence through corpus machine translation. And still is. As Kurt Sutter (show runner for Sons of Anarchy) taught us about Google, “[t]he truth is, they don’t give a shit about free speech, and are the antithesis of their own mantra, ‘Don’t be evil.’” That was 2014 and boy was he right. And he still is. It’s not just Google, but Google is emblematic of Silicon Valley.
One of the lessons we learned from the 1990s is the calvary is not coming. We have to take our own steps to work both cooperatively and defensively against a tech threat. The Human Artistry Campaign and its AI Principles effort is a hopeful indicator that the creative community and its partners are coming together to get ahead of both the threat and the promise of AI.
Let’s not forget that it’s not just about us, it’s also about the fan, our “consumers” if you will. The biggest threat to creators in my view is destroying the relationship of trust that exists between fans and creators. If AI can allow a machine to impersonate a creator, that deception harms the creator, surely. But it also harms the fan.
One of the AI principles from the Human Artistry Campaign jumped out at me as addressing this vital issue:
Trustworthiness and transparency are essential to the success of AI and protection of creators.
Complete recordkeeping of copyrighted works, performances, and likenesses, including the way in which they were used to develop and train any AI system, is essential. Algorithmic transparency and clear identification of a work’s provenance are foundational to AI trustworthiness. Stakeholders should work collaboratively to develop standards for technologies that identify the input used to create AI-generated output. In addition to obtaining appropriate licenses, content generated solely by AI should be labeled describing all inputs and methodology used to create it — informing consumer choices, and protecting creators and rightsholders.
Informing consumer choices.
For a moment forget the artistic integrity, forget the human intervention, forget the free riding, just for a moment because these are all vital issues, too. At the core of the AI problem is deception and that issue is as old as time. You can’t essentially deceive fans about the origin of a work and you certainly can’t build a machine that does this all the livelong day and pretend you didn’t.
In Book 2 of Plato’s Republic, he uses the legend of a magic ring that turns the bearer invisible to illustrate a dialog on the nature of justice. The ring turns the wearer invisible so that they are capable of doing all manner of things while invisible–or anonymous–that would clearly be both unjust and punishable without the ring. Plato asks if an act is unjust solely because you get caught or is it unjust regardless of whether you are hidden from sight or apprehension. Yep, those Greeks were onto this early.
Deception is not genius. At the core of our concerns about AI is keeping them honest to protect our fans and the bedrock of the creator-fan relationship. Consumers should be able to rely on the reality of what appears to be an artist’s work that it actually does come from that artist.
We do this with almost any other product or service that is placed into commerce, so why not with creative works? After all, artist rights are human rights.
We were happy to endorse the AI principles and encourage you to find out more about it at the Human Artistry Campaign or Artist Rights Watch and sign the petition.
