WesRecs x WHALE: Art Media Spotlight #007 (Generative AI / ChatGPT Edition)
A quick rundown of cool art reads, views, & listens curated for WHALE Members by Wes Hazard
A recurring series highlighting worthwhile writing, podcasts, & videos of note pertaining to digital art, NFTs, blockchain and WHALE Member’s mission to become the MoMA of the Metaverse.
With this edition of WesRecs x WHALE I wanted to focus on generative Artificial Intelligence, especially on content that has come out in the wake of OpenAI’s release of the publicly available ChatGPT client. In the last several weeks this release has simply dominated the tech discourse both inside and outside of crypto and has gone beyond that to reach the the popular consciousness by way of memes, breathless news coverage, and late night show monologues. As such it seemed only fitting to dig into some of the more interesting content I’ve seen. Trust me, this isn’t a tenth of what I’ve reviewed on the topic and I’ll doubtless have another generative AI-centric WesRecs in the not too distant future. In the meantime I hope you enjoy!
📜 Table of Recs (Contents):
⭐ This Is Why Generative Artificial Intelligence Is About To Change Humanity with Nina Schick — World Renowned Synthetic Media & Deepfake Expert
Welcome To The Metaverse Podcast [Dec 2022]
Welcome to the Metaverse, hosted by Luke Franks, is a podcast that not only endeavors to explore every nook & cranny of the emerging tech & culture related to its titular subject, it’s also one that consistently manages to deliver incredible alpha by profiling the major players and projects in the space before they blow up in the wider consciousness. Franks is a curious and thoughtful presenter who clearly spends considerable time both doing his research and building out his Rolodex of visionaries, founders, artists, & project heads to deliver a consistently informative and entertaining show where I’ve first heard about major metaversal players like RTFKT, Yield Guild Games, Teflon Sega, The Fabricant and many more.
The guest on this ep is author and cybersecurity/synthetic media/Generative AI expert Nina Schick whose 2020 book Deep Fakes: The Coming Infopocalypse has become an essential text in outlining the current landscape of potential future opportunities (and threats) of AI generated content as it relates to news, mass media, and (dis)information — although as Schick herself notes in the discussion, the blistering pace of tech development in the field has rendered certain parts of this publication obsolete already. Based on the vision and thoughtfulness heard from Schick in this discussion I will absolutely be diving into the book as soon as I’m able and I’d recommend the same to anyone curious about the dawning AI revolution.
I do think we’re entering a new era of human evolution where everything we have thus far believed to be unique to human creativity or intelligence, at least the production process, can now be done by AI and at scale so that’s going to have really profound implications and consequences.
I’m not sure if agree fully with that statement (mainly since a lot of human creativity has to do with meaning itself…which is exactly meaningless to AI) but when it comes to sheer scale and volume of output??? There’s just no question that AI will surpass us here and it’s not unreasonable to assume, as Schick does, that up to 90% of all digital content will be AI-produced by 2025. In fact, the angle that I found most interesting in this discussion is the one that argues that this level of AI production is necessary if the metaverse that so many of us imagine is to actually come to fruition. Just think, if we’re all going to spend significant time playing, socializing, shopping, and doing business in a vast and customizable metaverse we are going to need a LOT of content to fill it with and there are simply not enough devs & 3D modelers on planet Earth to fill that need. If we want a metaverse AI is going to be designing a lot of the clothing, wallpaper, and artifacts in it.
… this level of AI production is necessary if the metaverse that so many of us imagine is to actually come to fruition.
⭐ The College Essay Is Dead
Stephen Marche
The Atlantic [2022]
In the wake of OpenAI’s public release of ChatGPT there has been an endless wash of news articles, parodies, and think pieces (of both the laudatory & hand-wringing variety). Many of them have been in response to this specific piece from Stephen Marche in The Atlantic with its attention-grabbing title. From what I’ve seen personally, most of the responses have been rebuttals. Yet regardless of what you think of Marche’s thesis his essay is worth the read in terms of understanding how the new tech is being framed and evaluated by a great many people who work with language, learning, and education.
The essay, in particular the undergraduate essay, has been the center of humanistic pedagogy for generations. It is the way we teach children how to research, think, and write. That entire tradition is about to be disrupted from the ground up. Kevin Bryan, an associate professor at the University of Toronto, tweeted in astonishment about OpenAI’s new chatbot last week: “You can no longer give take-home exams/homework … Even on specific questions that involve combining knowledge across domains, the OpenAI chat is frankly better than the average MBA at this point. It is frankly amazing.” Neither the engineers building the linguistic tech nor the educators who will encounter the resulting language are prepared for the fallout.
Marche argues that while the tech is still young, it is already able to produce basic generative essays in response to prompts that equal or surpass the coherence and writing quality of a great number of his students. Since the essay is the primary tool that is used to assess the understanding & ability of most students in the humanities, it will soon be a meaningless gesture to assign such work since any given student could simply prompt a chat bot to come up with a passable (and passing) response in a matter of seconds. He is NOT arguing that this would demonstrate the student’s actual ability or knowledge in their given field, just that that in the system that currently exists they could put in next to no effort and “earn” the same credentials as a student that actually works to develop mastery.
Rather than simply throwing up his hands at this perceived eventuality Marche instead argues for an evolved understanding of what it is we’re actually trying to teach in the humanities (how to think) and puts forth the hope that these developments will lead to an increased cross-pollination of ideas between those in the humanities and the fields of computer science/data science/AI. Given the number of responses to this piece, Marche’s ideas have definitely struck a nerve, but given how nuanced his actual argument is it seems a lot of the people writing to rebut this stopped at reading the title instead of actually engaging.
The humanists will need to understand natural-language processing because it’s the future of language, but also because there is more than just the possibility of disruption here. Natural-language processing can throw light on a huge number of scholarly problems. It is going to clarify matters of attribution and literary dating that no system ever devised will approach; the parameters in large language models are much more sophisticated than the current systems used to determine which plays Shakespeare wrote, for example. It may even allow for certain types of restorations, filling the gaps in damaged texts by means of text-prediction models. It will reformulate questions of literary style and philology; if you can teach a machine to write like Samuel Taylor Coleridge, that machine must be able to inform you, in some way, about how Samuel Taylor Coleridge wrote.
⭐ ChatGPT Is Dumber Than You Think
Ian Bogost
The Atlantic [2022]
OK, not gonna lie: I kind of hated this piece at first. It seemed to me that Bogost was judging a newly-public tech far too harshly in its first week of release. I was flabbergasted that he did not perceive the very likely probability that we’d see exponential growth in the quality and finesse of the generative text outputs within a few years (or much sooner). But I kept going and by the end of this I was actually kind of in love with his stance. I now view it as one of the most nuanced takes on the limitations, and limitless creative possibilities, of the new frontier of AI generated text.
Essentially: this tech is amazing and awe-inspiring…within a very very narrow band that will never surpass human ability, but which can certainly help us to expand the playing area of that ability.
Others, more measured but no less bewitched, may claim that “it’s still early days” for a technology a mere few years old but that can already generate reasonably good 12th-century lyric poems about Whataburger. But these are the sentiments of the IT-guy personalities who have most mucked up computational and online life, which is just to say life itself. OpenAI assumes that its work is fated to evolve into an artificial general intelligence — a machine that can do anything. Instead, we should adopt a less ambitious but more likely goal for ChatGPT and its successors: They offer an interface into the textual infinity of digitized life, an otherwise impenetrable space that few humans can use effectively in the present.
As has been reinforced over and over again by the most perceptive and hype-averse commentators writing about the ChatGPT wave:
The AI knows NOTHING.
The prompts you submit and the outputs it responds have no meaning or value to it. It does not feel, it does not care, it is neither creative nor smart. It is, however, a ridiculously powerful tool capable of helping humans, who do possess those qualities, to explore and probe the limits and depths of language and knowledge in a new way. ChatGPT and similar AI innovations deserve serious hype, but of a very specific kind.
⭐ I tried ChatGPT from OpenAI and my mind was blown
Linh Dao Smooke
Hacker Noon [12/1/22]
This wasn’t the most substantive piece of content out there but it did include this, and honestly this right here is one of the best examples I’ve seen of the sort of simple, everyday, unsexy task that these generative text technologies can help humans with immediately. Human creativity will be safe no matter what technologies we develop but if this stuff can get us out of the drudgery of customer service emails I am 100% for it.
⭐ Generative AI needs Blockchain
Andrew Christian Davis (Medium)
I think that the underlying tech of crypto & blockchains represent a monumental evolution of the promise and possibility of the Internet. I believe that NFTs, on on a basic utility level, will be integrated into nearly every facet of our digital existence and that they will see such widespread adoption as to be as routine as email, SMS, & hyperlinks. That said, I also believe that the continued development and spread of artificial intelligence will, as a class of tech, have a cultural/political/financial impact that is orders of magnitude greater and more obvious than even that of blockchains. AI is a very very big deal that I don’t think we’ve begun to reckon with the smallest fraction of what that will mean.
Blockchain-based data and IP infrastructure and protocols could underpin this entire system at every stage in the AI value chain. This means that every participant in this ecosystem — those who provide data, train models, create applications, and produce IP from those applications — can all benefit and participate in the upside. Not only would every participant layer receive compensation or some meaningful value from the layer above (ex. model developers paying for data, application developers paying model developers for access to their models, etc.), value could flow and accrue down the whole stack as end users purchase and engage with content/data.
If these two technologies are to be as widespread and meaningful as it seems they will be, it is inevitable that they will intersect and overlap and merge. And while I won’t pretend to any ability to fully grasp what that will mean, one obvious problem in AI seems to have a ready made solution in blockchains.
Namely 👉 authenticating data.
We simply need to be able to verify, track, and properly attribute/license the sources of the massive amounts of data used to train models like ChatGPT and and we need to be able to do the same for the endless outputs that they will produce. This is essential for compensating creatives, analyzing bias/intent, battling plagiarism & disinformation, refining future versions of AI and so much more. If we’re going to spend more and more of our time in digital environments and if a huge portion of those environments is to be generated by AI it is imperative that we have accurate info about how those environments were made, and by what tools, and who initiated their creation. Immutable & verifiable blockchain tech seems like the perfect solution to this problem, as this piece illustrates. It’s hardly the application with the greatest “wow” factor but it seems like the perfect match.
We’re in the middle of a light appetizer in a generative AI meal and it’s exciting to see what’s ahead. It’s also critical to remember the past. We’ve had a bitter taste of what happens when we allow tech to progress without safeguards and considering the costs. The stakes will be higher this time.
OK, those are the current recs, but certainly not the only AI & ART-related media that we’re discussing in WHALE Members!
Be sure to stop by our Discord and check out the 👉 #art-edu 👈 channel for additional recs & dope conversation!
🐳 WHALE MEMBERS RESOURCES
⚫ ️Website: https://whale.me/
⚫ ️Discord: https://discord.gg/whale
⚫️ Twitter: https://twitter.com/WHALEMembers
⚫️ YouTube: https://www.youtube.com/whalemembers
⚫️ Instagram: https://www.instagram.com/whalemembers
⚫️ Twitch: https://www.twitch.tv/whalemembers
⚫️ Reddit: https://www.reddit.com/r/WHALEMembers
⚫️ Medium: https://medium.com/whale-members
⚫️ Telegram: https://t.me/WHALEMembers
