AI and the Future of Education
Should Education be Exempt from LLM Copyright Restrictions?
Image source — Rawpixel.com
This morning Zanny Minton Beddoes, editor-in-chief of the Economist, interviewed OpenAI’s Sam Altman and Microsoft’s Satya Nadella. I connected to the interview by Zoom mainly because I was interested to see how these tech industry titans would address one of the big questions about the future of AI Large Language Models, i.e. how are the original content creators going to be compensated for what is in effect a sophisticated and high-speed tool for repackaging and regurgitating the work of others. I’m sure I was one of hundreds who submitted that question in advance. However, Ms. Beddoes chose not to include that question and mainly stuck to softball questions about next-gen enhancements.
I don’t know if Sam and Satya believe that the outputs of Chat GPT 4 and various Microsoft tools constitute derivative works. As Google’s AI tells me — “A derivative work must contain some protectable expression that is not found in a prior work. It must also contain sufficient protectable expression from a prior work such that distributing it without the permission of the author (or rights holder) of the prior work would violate their copy.” I guess that’s a question for the lawyers. I don’t kid myself that my smatterings of literary works have even made the input list for LLMs but if they ever did, I think I would want some commercial compensation for the use of my intellectual property. With one possible exception, the field of education.
Mr. Nadella spoke this morning of a future where all children have cost-effective access to their own AI tutor. I have spent years bemoaning how the internet has increased ignorance by filtering and shielding us from any facts that do not align with our pre-existing world view. Is it possible that AI tutors could foster critical thinking and somehow circumvent the unquestioning absorption of unsubstantiated rumor that drives the commercial success of social media platforms? I’m not sure. Chat GPT 4 already suffers from the risks of garbage in — garbage out. Like many people I played with it when it first came out and I found its vague generalizations and occasional inaccuracies to be uninspiring. If we legally tie up access to thorough and substantiated analysis, will Chat GPT 5 end up being full of nonsense conspiracy theories scraped from lunatic rants online? Clearly that’s not the tutor Mr. Nadella is envisioning.
I don’t know the solution. Most education budgets around the world have the girth of a shoestring and layering an expensive financial compensation model on access to a factual and unbiased AI tutor is not in society’s best interest. It’s my hope that somehow, we can give children access to the tools they need to think for themselves and protect the copyright of the content creators.
