17 October, 2024
The case against AI art
No matter who powerful generative AI becomes, writer Ted Chiang says it will never create true art.
Chiang is one of the most admired science fiction authors writing today, best known for the novella “Story Of Your Life” (which was adapted into the movie “Arrival”). But he’s also published terrific pieces for The New Yorker looking at the dangers and shortcomings of AI.
You should really read his latest article in its entirety, but briefly: Chiang argues that the potential of large language models remains “largely theoretical” — thus far, generative AI has been most successful at “lowering our expectations, both of the things we read and of ourselves when we write anything for others to read. It is a fundamentally dehumanizing technology because it treats us as less than what we are: creators and apprehenders of meaning.”
Even as LLMs improve, Chiang argues that their output will never be art — which he acknowledges is “notoriously hard to define,” but he tries anyway: “Art is something that results from making a lot of choices.” Sure, those choices might not result in a particularly good novel or painting or film, but you’re still “engaged in an act of communication between you and your audience.”
“We are all products of what has come before us, but it’s by living our lives in interaction with others that we bring meaning into the world,” Chiang concludes. “That is something that an auto-complete algorithm can never do, and don’t let anyone tell you otherwise.”
Redfin is already trying to defend against a new flat-fee real estate startup
Redfin is responding to a new startup that is hoping to upend the way people search for and buy homes by offering a flat-fee service.
On August 29, TechCrunch reported that a startup called Landian had emerged from stealth to offer homebuyers a way to tour and make offers on homes through a flat-fee service, rather than paying commissions.
That company was co-founded by Josh Sitzer, who sued the National Association of Realtors (NAR) in a landmark case over agent commissions. Under the resulting settlement, the NAR agreed to pay $418 million in damages and to abolish the Participation Rule, which required sell-side agents to make an offer of compensation to buyer brokers. That and other rule changes are expected to transform the real estate market.
Redfin is skeptical about the flat-fee model, although it described Landian as “a brother in arms, eager like us to give consumers a better deal.” The 18-year-old company once tried a similar model, and explains why it didn’t work:
“When we tried this before in a fiercely competitive housing market, we struggled to win on behalf of customers the offer-writing agent hadn’t met, for listings that agent hadn’t seen,” a spokesperson said. “We also learned that when customers want to call on the expertise of one person, morning, noon, and night, you have to pay that person very, very well. For now, we believe we can offer homebuyers the best value by using Redfin.com to eliminate the single largest cost of being an agent, which is finding customers, and by pairing the industry’s best agents with lending and title services.”
Redfin points out that it charges commissions as low as 1% to home sellers and as low as 2% to homebuyers, and claims to have saved its customers $1.6 billion in fees.
“Unlike Landian, we don’t charge for tours or require customers to hire an agent sight unseen,” a spokesperson said.
Redfin went on to say that it “may experiment again” with a flat-fee itemized service. But it’s wary.
The RIAA's lawsuit against generative music startups will be the bloodbath AI needs
Like many AI companies, music generation startups Udio and Suno appear to have relied on unauthorized scrapes of copyrighted works in order to train their models. This is by their own and investors’ admission, as well as according to new lawsuits filed against them by music companies. If these suits go before a jury, the trial could be both a damaging exposé and a highly useful precedent for similarly sticky-fingered AI companies facing certain legal peril.
The lawsuits, filed by the Recording Industry Association of America (RIAA), put us all in the uncomfortable position of rooting for the RIAA, which for decades has been the bogeyman of digital media. I myself have received nastygrams from them! The case is simply that clear.
The gist of the two lawsuits, which are extremely similar in content, is that Suno and Udio (strictly speaking, Uncharted Labs doing business as Udio) indiscriminately pillaged more or less the entire history of recorded music to form datasets, which they then used to train a music-generating AI.
And here let us quickly note that these AIs don’t “generate” so much as match the user’s prompt to patterns from their training data and then attempt to complete that pattern. In a way, all these models do is perform covers or mashups of the songs they ingested.
That Suno and Udio did ingest said copyrighted data seems, for all intents and purposes (including legal ones), very likely. The companies’ leadership and investors have been unwisely loose-lipped about the copyright challenges of the space.
They have admitted that the only way to create a good music generation model is to ingest a large amount of high-quality music. It is very simply a necessary step for creating machine learning models of this type.
Then they said that they did so without the permission of music labels. Investor Antonio Rodriguez of Matrix Partners told Rolling Stone just a few months ago:
Honestly, if we had deals with labels when this company got started, I probably wouldn’t have invested in it. I think that they needed to make this product without the constraints.
The companies told the RIAA’s lawyers that they believe the media it has ingested falls under fair-use doctrine — which fundamentally only comes into play in the unauthorized use of a work. Now, fair use is admittedly a complex and hazy concept in idea and execution, but the companies’ use does appear to stray somewhat outside the intended safe harbor of, say, a seventh grader using a Pearl Jam song in the background of their video on global warming.
To be blunt, it looks like these companies’ goose is cooked. They might have hoped that they could take a page from OpenAI’s playbook, using evasive language and misdirection to stall their less deep-pocketed critics, like authors and journalists. (If by the time AI companies’ skulduggery is revealed and they’re the only option for distribution, it no longer matters.)
But it’s harder to pull off when there’s a smoking gun in your hand. And unfortunately for Udio and Suno, the RIAA says in its lawsuit that it has a few thousand smoking guns and that songs it owns are clearly being regurgitated by the music models. Its claim: that whether Jackson 5 or Maroon 5, the “generated” songs are lightly garbled versions of the originals — something that would be impossible if the original were not included in the training data.
The nature of LLMs — specifically, their tendency to hallucinate and lose the plot the more they write — precludes regurgitation of, for example, entire books. This has likely mooted a lawsuit by authors against OpenAI, since the latter can plausibly claim the snippets its model does quote were grabbed from reviews, first pages available online and so on. (The latest goalpost move is that they did use copyright works early on but have since stopped, which is funny because it’s like saying you only juiced the orange once but have since stopped.)
What you can’t do is plausibly claim that your music generator only heard a few bars of “Great Balls of Fire” and somehow managed to spit out the rest word for word and chord for chord. Any judge or jury would laugh in your face, and with luck a court artist will have their chance at illustrating that.
The current legal cases against generative AI are just the beginning
This is not only intuitively obvious but legally consequential as well, as the re-creation of entire works (garbled, but quite obviously based on the originals) opens up a new avenue for relief. If the RIAA can convince the judge that Udio and Suno are doing real and major harm to the business of the copyright holders and artists, it can ask the court to shut down the AI companies’ whole operation at the outset of the trial with an injunction.
Opening paragraphs of your book coming out of an LLM? That’s an intellectual issue to be discussed at length. Dollar-store “Call Me Maybe” generated on demand? Shut it down. I’m not saying it’s right, but it’s likely.
The predictable response from the companies has been that the system is not intended to replicate copyrighted works: a desperate, naked attempt to offload liability onto users under Section 230 safe harbor. That is, the same way Instagram isn’t liable if you use a copyrighted song to back your Reel. Here, the argument seems unlikely to gain traction, partly because of the aforementioned admissions that the company itself ignored copyright to begin with.
What will be the consequence of these lawsuits? As with all things AI, it’s quite impossible to say ahead of time, since there is little in the way of precedent or applicable, settled doctrine.
My prediction is that the companies will be forced to expose their training data and methods, these things being of clear evidentiary interest. And if this evidence shows that they are indeed misusing copyrighted material, we’ll see an attempt to settle or avoid trial, and/or a speedy judgment against Udio and Suno. It’s likely that at least one of the two will attempt to continue onward, using legal (or at least legal-adjacent) sources of music, but the resulting model would (by their own standards for training data) almost certainly result in a huge step down in quality, and users would flee.
Investors? Ideally, they’ll lose their shirts, having placed their bets on something that was in all likelihood illegal and certainly unethical, and not just in the eyes of nebbish author associations but according to the legal minds at the infamously and ruthlessly litigious RIAA.
The consequences may be far-reaching: If investors in a hot new generative media startup suddenly see a hundred million dollars vaporized due to the fundamental nature of generative media, suddenly a different level of diligence will seem appropriate.
Companies may learn from the trial or settlement documents what can be said — or perhaps more importantly, what should not be said — to avoid liability and keep copyright holders guessing.
Though this particular suit seems almost a foregone conclusion, it will not be a playbook to prosecuting or squeezing settlements out of other generative AI companies but an object lesson in hubris.
It’s good to have one of those every once in a while, even if the teacher happens to be the RIAA.
Ethereum co-founder's warning against 'pro-crypto' candidates: 'Are they in it for the right reasons?'
Vitalik Buterin, the co-founder of Ethereum, issued a warning on Wednesday against choosing a candidate purely based on whether they claim to be “pro-crypto.” In a blog post, Buterin said it’s more important to scrutinize a candidate’s broader policies to ensure they support cryptocurrency’s underlying goals, including internationalism and protection for private communications.
“If a politician is pro-crypto, the key question to ask is: Are they in it for the right reasons?,” wrote Buterin. “Do they have a vision of how technology and politics and the economy should go in the 21st century that aligns with yours?”
Though Buterin does not mention any politicians or crypto investors by name, his comments come just one day after Marc Andreessen and Ben Horowitz threw their support behind former President Donald Trump in the 2024 Presidential election. The founders of Andreessen Horowitz noted on their podcast yesterday that Trump’s crypto regulation plan is “a flat-out blanket endorsement of the entire space.” The influential venture capitalists join the ranks of other notable Silicon Valley players, including Elon Musk, who endorsed Trump in the last week.
Further, Ethereum’s co-founder made the case that signaling you broadly support any “pro-crypto” candidates could incentivize politicians to promote the cause in bad faith. Buterin notes that authoritarian leaders, particularly in Russia, have claimed to support crypto in an effort to consolidate power.
“It doesn’t matter if they also support banning encrypted messaging, if they are a power-seeking narcissist, or if they push for bills that make it even harder for your Chinese or Indian friend to attend the next crypto conference — all that politicians have to do is make sure it’s easy for you to trade coins,” said Buterin.
The co-founder of Ethereum suggested to look into a “crypto-friendly” politician’s views on crypto five years ago. He says this can serve as a guide for whether the politician may reverse their position five years in the future.
Notably, former President Trump starkly opposed decentralized tokens five years ago. In a tweet from July 2019, Trump said he’s “not a fan of Bitcoin and other Cryptocurrencies, which are not money.” In a follow-up tweet, he said “we have only one real currency in the USA,” referring to the United States dollar.
But in May, Trump completed a total flip-flop on his stance regarding cryptocurrencies, becoming the first major presidential candidate to accept bitcoin donations. The Wall Street Journal reports that Trump’s crypto fundraising efforts have collected $3 million worth of donations in the second quarter.
'Model collapse': Scientists warn against letting AI eat its own tail
When you see the mythical Ouroboros, it’s perfectly logical to think, “Well, that won’t last.” A potent symbol — swallowing your own tail — but difficult in practice. It may be the case for AI as well, which, according to a new study, may be at risk of “model collapse” after a few rounds of being trained on data it generated itself.
In a paper published in Nature, British and Canadian researchers led by Ilia Shumailov at Oxford show that today’s machine learning models are fundamentally vulnerable to a syndrome they call “model collapse.” As they write in the paper’s introduction:
We discover that indiscriminately learning from data produced by other models causes “model collapse” — a degenerative process whereby, over time, models forget the true underlying data distribution …
How does this happen, and why? The process is actually quite easy to understand.
AI models are pattern-matching systems at heart: They learn patterns in their training data, then match prompts to those patterns, filling in the most likely next dots on the line. Whether you ask, “What’s a good snickerdoodle recipe?” or “List the U.S. presidents in order of age at inauguration,” the model is basically just returning the most likely continuation of that series of words. (It’s different for image generators, but similar in many ways.)
But the thing is, models gravitate toward the most common output. It won’t give you a controversial snickerdoodle recipe but the most popular, ordinary one. And if you ask an image generator to make a picture of a dog, it won’t give you a rare breed it only saw two pictures of in its training data; you’ll probably get a golden retriever or a Lab.
Now, combine these two things with the fact that the web is being overrun by AI-generated content and that new AI models are likely to be ingesting and training on that content. That means they’re going to see a lot of goldens!
And once they’ve trained on this proliferation of goldens (or middle-of-the road blogspam, or fake faces, or generated songs), that is their new ground truth. They will think that 90% of dogs really are goldens, and therefore when asked to generate a dog, they will raise the proportion of goldens even higher — until they basically have lost track of what dogs are at all.
This wonderful illustration from Nature’s accompanying commentary article shows the process visually:
A similar thing happens with language models and others that, essentially, favor the most common data in their training set for answers — which, to be clear, is usually the right thing to do. It’s not really a problem until it meets up with the ocean of chum that is the public web right now.
Basically, if the models continue eating each other’s data, perhaps without even knowing it, they’ll progressively get weirder and dumber until they collapse. The researchers provide numerous examples and mitigation methods, but they go so far as to call model collapse “inevitable,” at least in theory.
Though it may not play out as the experiments they ran show it, the possibility should scare anyone in the AI space. Diversity and depth of training data is increasingly considered the single most important factor in the quality of a model. If you run out of data, but generating more risks model collapse, does that fundamentally limit today’s AI? If it does begin to happen, how will we know? And is there anything we can do to forestall or mitigate the problem?
The answer to the last question at least is probably yes, although that should not alleviate our concerns.
Qualitative and quantitative benchmarks of data sourcing and variety would help, but we’re far from standardizing those. Watermarks of AI-generated data would help other AIs avoid it, but so far no one has found a suitable way to mark imagery that way (well … I did).
In fact, companies may be disincentivized from sharing this kind of information, and instead hoard all the hyper-valuable original and human-generated data they can, retaining what Shumailov et al. call their “first mover advantage.”
[Model collapse] must be taken seriously if we are to sustain the benefits of training from large-scale data scraped from the web. Indeed, the value of data collected about genuine human interactions with systems will be increasingly valuable in the presence of LLM-generated content in data crawled from the Internet.
… [I]t may become increasingly difficult to train newer versions of LLMs without access to data that were crawled from the Internet before the mass adoption of the technology or direct access to data generated by humans at scale.
Add it to the pile of potentially catastrophic challenges for AI models — and arguments against today’s methods producing tomorrow’s superintelligence.
X files antitrust suit against advertising groups over ‘systematic illegal boycott’
X CEO Linda Yaccarino on Tuesday announced that the social media platform has filed an antitrust lawsuit against the Global Alliance for Responsible Media (GARM) and the World Federation of Advertisers (WFA).
In a video posted to X, Yaccarino accuses the organizations — along with GARM members CVS Health, Mars, Orsted and Unilever — of what Yaccarino calls a “systematic illegal boycott” of the platform.
A Message to X Users pic.twitter.com/6bZOYPhWVa
— Linda Yaccarino (@lindayaX) August 6, 2024
The executive cites a July report from the U.S. House of Representatives Judiciary Committee titled, “GARM’s (Global Alliance for Responsible Media) Harm.” According to the report:
Through GARM, large corporations, advertising agencies, and industry associations participated in boycotts and other coordinated action to demonetize platforms, podcasts, news outlets, and other content deemed disfavored by GARM and its members. This collusion can have the effect of eliminating a variety of content and viewpoints available to consumers.
GARM was founded by the World Federation of Advertisers in 2019 in a bid to “help the industry address the challenge of illegal or harmful content on digital media platforms and its monetization via advertising,” according to the organization’s site.
The Judiciary report specifically addresses boycotts of X, The Joe Rogan Experience/Spotify and “Candidates, platforms, and news outlets with opposing political views.”
In particular it addresses organization member concerns over Elon Musk’s acquisition of the platform then known as Twitter. One member, according to the report, suggested that fellow members stop paid advertisements on the service, contributing to a precipitous drop in revenue.
“GARM’s internal documents show that GARM was asked by a member to ‘arrange a meeting and hear more about [GARM’s] perspectives about the Twitter situation and a possible boycott from many companies,” the report’s authors note. GARM also held ‘extensive debriefing and discussion around Elon Musks’ [sic] takeover of Twitter,’ providing ample opportunity for the boycott to be organized.”
In her own statement, Yaccarino claims that GARM’s “illegal behavior of these organizations and their executives cost X billions of dollars.”
We tried being nice for 2 years and got nothing but empty words.
Now, it is war. https://t.co/MEzH0vqz0p
— Elon Musk (@elonmusk) August 6, 2024
Musk was less measured in his response, posting, “We tried being nice for 2 years and got nothing but empty words. Now, it is war.” The executive had similarly incendiary words for advertisers last year, stating, “If somebody’s going to try to blackmail me with advertising, blackmail me with money? Go f*** yourself. Go. F***. Yourself. Is that clear?”
He also promised at the time to document companies participating in the boycott “in great detail.”
The suit follows a recent governmental crackdown on tech antitrust. Yesterday, Google lost a landmark battle alleging the software giant of maintaining a search monopoly through illegal acts.
X joined GARM in early July, noting, “X is committed to the safety of our global town square and proud to be part of the GARM community” through its Safety account.
Artists' lawsuit against generative AI makers can go forward, judge says
A class action lawsuit filed by artists who allege that Stability, Runway and DeviantArt illegally trained their AIs on copyrighted works can move forward, but only in part, the presiding judge decided on Monday. In a mixed ruling, several of the plaintiffs’ claims were dismissed while others survived, meaning the suit could end up at trial. That’s bad news for the AI makers: Even if they win, it’s a costly, drawn-out process where a lot of dirty laundry will be put on display. And they aren’t the only companies fighting off copyright claims — not by a long shot.
UAW files federal labor charges against Trump, Musk for intimidating workers at X Spaces event
The United Auto Workers union said Tuesday that it filed federal labor charges against Donald Trump and Elon Musk.
The union alleges that Trump and Musk attempted to “threaten and intimidate workers” who engage in strikes after comments made during the now notorious X Spaces interview Monday evening.
“They go on strike, I won’t mention the name of the company, but they go on strike and you say, That’s OK, you’re all gone. You’re all gone. So, every one of you is gone,” Trump told Musk.
The UAW — which recently endorsed Vice President Kamala Harris for president — says those comments can be construed as advocating for the firing of striking workers, which is illegal under the National Labor Relations Act.
“Both Trump and Musk want working class people to sit down and shut up, and they laugh about it openly. It’s disgusting, illegal, and totally predictable from these two clowns,” said UAW president Shawn Fain in a statement.
UAW files federal labor charges against Trump, Musk for intimidating workers at X Spaces event
The United Auto Workers union said Tuesday that it filed federal labor charges against Donald Trump and Elon Musk.
The union alleges that Trump and Musk attempted to “threaten and intimidate workers” who engage in strikes after comments made during the now notorious X Spaces interview Monday evening.
“They go on strike, I won’t mention the name of the company, but they go on strike and you say, That’s OK, you’re all gone. You’re all gone. So, every one of you is gone,” Trump told Musk.
The UAW – which recently endorsed Vice President Kamala Harris for president – says those comments can be construed as advocating for the firing of striking workers, which is illegal under the National Labor Relations Act.
“Both Trump and Musk want working class people to sit down and shut up, and they laugh about it openly. It’s disgusting, illegal, and totally predictable from these two clowns,” said UAW president Shawn Fain in a statement.