LLMs — and, more broadly, generative AIs — are having some surprising secondary effects, and these are coming far more quickly than its promoters might have expected.
These second-order effects are as profound as those of the initial Internet Age, which seem to still be slowly emerging. Comparatively, society and stakeholders are not sitting idly by this time, waiting to see what transpires or hoping for the best. Instead, many are actively engaged in constraining, restraining, and refining possible paths forward for LLMs and the companies rolling them out.
Here are some of the second-order effects I’ve seen.
A Re-embrace of IP Laws and Rights
Musicians, publishers, authors, artists, and photographers are actively engaged in ensuring that IP rights (copyright, trademarks, moral rights) are honored, protected, and enumerated in LLMs, requiring licenses, payments, and disclosures from the companies using legally protected content.
This is in stark contrast to 20-25 years ago, when the habit of mindlessly feeding content to newfangled intermediaries was the norm. Even those sins are being revisited, with a raft of lawsuits and regulations coming at technology companies to rectify past misdeed and missed payments as journalists and others realize they’ve been had.
LLMs have shocked awake the stirring IP giant.
Regulation and Legislation
Governments have been swift to hold hearings, draft legislation, and pass rules for LLM and generative AI companies, with the EU taking the first major step last week. It won’t be the last, as the US is poised to move forward soon, as well. Only China is an outlier, for political reasons that under the current regime are concerning.