Individual priority and credit in the age of AI: When the question IS the solution, our queries help AI learn and our IP gets scraped into the ethersphere
By Terry Poulos
(NOTE: This article was published first on Medium minutes prior to appearing on Scientiquity.com
Objection, your honor. Leading the witness. It’s not solely a page out of a courtroom drama. It’s happening every time you ask a question of artificial intelligence (AI). You direct the machine toward the answer. This is the modern-day equivalent of the Riddle of the Sphinx.
In a way, to ask is to give. Careful what you ask for.
AI requires reams of data to scrape and subsume into its memory banks, in order for it to learn and improve. Evolve, so to speak. But will AI one day “kill” priority of creation and individual ownership of intellectual property? Notwithstanding patents. Does the individual still matter? Is the lone genius destined to have the fruit of their lifetime go unnoticed, unappreciated, unrewarded? Stolen away, repackaged and regifted?
I ask these questions because creators warrant recognition for their efforts. We all have egos and for many it’s what drives us to reach for greatness. But that’s just the half of it. Why should anyone put years of hard work into something, thousands of hours of research, experimentation and refinement, proprietary intuition and knowledge, with all the time and mental strain and vexing effort…only to see AI run with the output by instituting a slight tweak and spin on it, then giving it away to the next AI user who then usurps the original and authentic source?
And the cycle repeats ad infinitum.
Is this a product of singularity? The death of individual credit for the good of the collective?
If the next Einstein were to upload a complete Theory of Quantum Gravity (or Quantum Relativity), to the internet, replete with equations and fully-vetted philosophical and scientific formalism, would AI give credit where credit is due, even after AI churns through it and spits out an improved version? That is coming. Or the next user stumbles along and uses your queries and IP to take your theory to yet new heights.
Will the creative spark, the initial impetus for this world-changing paradigm, be recognized and rewarded? Or does it all get lost in the dustBIT of zeros and ones?
Granted, a community of humans — most notably the academic community — will be the ultimate arbiter of such a profound new paradigm. But the same way that computers are now “proofing” mathematics theorems and hypotheses, one would have to think AI will, in the end, play a substantial role in meting out the “last man (or bot) standing.” Incidentally, many mathematicians do not believe a computer proof is any proof at all.
In any event, is AI egalitarian? Does it have ethics? Can it acquire them? Does AI even possess the wherewithal to be the ultimate arbiter of truth?
We know AI has or will achieve critical mass. Eventually, large language models will feed on other large language models and intelligence will converge to a kind of singularity. The machine (collective world wide web) will become the overriding authority, and a hoard of facts and reality will absolutely be lost in the shuffle.
Priority and credit swept away in the dustBIT of history.
That said, should priority and assigning credit be the sole domain of humans? The folks at Cornell University maintain the not-for-profit global database ArXiv.org, the online e-print repository of all (quote) serious academic papers. However, you can’t get in the club if you’re not affiliated with a major institution. You must hold a Ph.D or Masters degree which is tethered to an established, recognized institution of higher learning. Otherwise, you can’t publish there (it’s a bit more complex, but that’s the general picture).
Of course, there’s a vast wasteland of papers and postulates and you have to draw the line somewhere, despite the fact Ph.D’s are not immune to publishing nonsensical waste product. Restrictions or no restrictions, ArXiv is not the end-all of facts, reality and priority.
Your other options, as a dilettante or novice, are to publish on Medium or some other portal and pray the platform lives in perpetuity, and is accessible to those who can effect change and assign priority.
Personally, I’m leaning toward a smart contract on the blockchain. It’s time-stamped and a permanent record, it’s verified across the seven-node network by mining (community consensus) or proof of work. The chain is strong and contrary to fear-mongering, as secure or more than traditional portals. It will become yet another “killer app” use case for academia and knowledge, as it has for currencies and now real estate, medical records, and so much more.
But can anyone find the paper efficiently, cite it and link to it? Today, only five-percent of the population in America is on the chain (note: That WILL change radically). Most importantly, can the scientific community find it?
The chain might be the solution for a future Michael Faraday. An apprentice with no formal education, the 19th century scientific prodigy is responsible for giving us the key insights into electromagnetic fields, which directly led to James Clerk Maxwell publishing the formal equations of EM, and that butterflied to Einstein developing Relativistic equations for the gravitational field.
In today’s physics, fields are everything. Faraday is a visionary who bulldozed the path for modern physics. He’s a giant of science worthy of mention with the likes of Maxwell, Niels Bohr, Albert Einstein, and Paul Durac.
Would Faraday get credit today, or be usurped by AI? He had to fight for his priority in the 19th century, and that was before the internet could — to paraphrase Mark Twain — “allow a lie to traverse the metaverse before the truth can get its boots on.”
There are two other potential outcomes. It’s also possible AI, in conjunction with a proof-of-work consensus blockchain, might make lying and deceit and misdirection and undermining of due priority less likely, or have a slightly indifferent influence.
Nonetheless, it is becoming a `Catch 22′ conundrum. Darned if you do ask AI questions, darned if you don’t. AI has gone mainstream and soon we won’t be able to compete unless we’re fully onboard. However, by doing so, we in part feed the beast. Teach the monster. Fuel the machine. Give it key clues!
We’re teaching AI new tricks. Human tricks. Biological creativity and uniqueness that AI can not yet itself produce. That could some day change. In the interim, AI is only as good as the input of us biological machines.
Questions — placed in context and in a certain sequence — can be as good as a solution. In the ethersphere, it’s `leading the witness” and there’s no judge to step in and play arbiter. AI may one day be the sole arbiter.
So be careful of what you ask AI. You may be giving away the store (advice: Protect yourself before going on AI to ask).
This all begs the question, will the give-and-take of AI eventually lead to the death of individual priority? Or will AI aid the greatness of individualism? Or a little of both. The truth is being written as fast as the speed of thought and brute force computation — ours and AI’s, respectively.
Terry Poulos is a writer, artist, mathematics and physics autodidact, and fractal geometer whose inquests include a Theory of Everything, archaeology and ancient technology and science, and more
Scientiquity 