Off the Charts
Music
What makes a song a hit? According to Max Martin, one of the industry’s most prolific and successful songwriters, the trick is to get to the ‘good stuff’ quickly. He believes - and an Economist analysis confirms - that the quicker you get through a song’s intro into the vocals, the better:
Another trend in music in the last twenty years is making a song’s overall length shorter:
A 2018 study by San Francisco-based engineer Michael Tauberg concluded that songs on the Billboard Hot 100 shed around 40 seconds since 2000, falling from 4:10-ish to roughly 3:30. The average length of the top 50 tracks on Billboard‘s year-end Hot 100 in 2021 was even less, a mere 3:07.
This can be attributed to the rise of streaming as a growing percentage of music delivery (interesting tidbit: it also happened in the 1960s as AM radio became popular). A shorter song is less skippable, and songs with less skips are prioritized in a streaming platform’s shuffle algorithm. Artists are often only paid royalties if their song streams for at least 30 seconds, encouraging them to create both shorter and catchier songs that dive straight into vocals and hooks.
Speaking of hooks, TikTok has thrown another wrinkle into the way songwriters operate - artists can blown up without even completing a song. In some cases, simply writing a catchy hook and waiting for it to go crazy on TikTok is enough:
Since 2020, if not before, a heap of young acts have gone viral with the former and then scrambled to transform them into the latter — to build a full track around the snippet that captivated TikTok.
They’re under pressure to quickly put out a track because they can’t earn much of anything streaming on TikTok:
A marketer who oversaw the campaign for a single that was used in roughly half a million TikTok videos reports that his artist took home less than $5,000 from the platform, despite the views numbering in the billions. TikTok’s parent company, ByteDance, “doesn’t view music as a value add,” says another senior executive. “They just view music as a cost center they have to limit as much as possible.”
Despite popular music driving the site’s meteoric growth, TikTok appears actively hostile to compensating those artists for their work. Even artists represented by labels are faced with a choice - sign a cut-rate deal with TikTok or spend resources chasing IP around the platform issuing takedowns. By allowing its users to upload videos with any music they want, TikTok is encouraging the unlicensed use of millions of songs, paying labels and creators essentially nothing.
Artists have come to view TikTok as a virality tool, using it to drive traffic to traditional streaming platforms like Spotify to make their money (we’ve talked before about Spotify’s payola schemes.) Rather than bending to industry pressure to be more transparent about its pay-to-play and in-house streaming suggestion lists, the company has doubled down on the practice with its Discovery Mode program:
Spotify officially announced that it has expanded access to Discovery Mode, a contentious program that gives artists the chance to gain more algorithmic exposure on the platform — through Spotify Radio and autoplay — in exchange for a lower royalty rate.
For even less royalties - fractions of a fraction of a penny - artists can potentially gain more listeners through Spotify’s algorithms. Algorithms that are entirely opaque, of course, and still put an artist’s songs at the mercy of skip rate and other unknown metrics.
So we’ve got artists writing shorter songs - or simply hooks they may turn into songs - to game streaming algorithms. Then those streaming platforms offer them the choice to forego more of their scant royalties to go viral on their platforms. Grim!
If you’re a popular enough artist, however, you can still rely on Billboard to boost virality the old fashioned way. Until recently, labels could game the charts by bundling albums with concert tickets or merchandise to pad sales numbers. With that avenue gone, some artists have turned to other methods to top the charts - writing a lot of songs:
Morgan Wallen’s “One Thing at a Time” is an 18-wheeler of an album, with 36 songs that run for a total of nearly two hours.
[…]
The thinking is simple: Fans listening to an album for the first time are likely to listen all the way through, so artists should give them more songs when they can.
Right. If you saw Wallen’s songs popping up on Billboard and wondered who on earth he is, the answer is: a young country artist who writes dozens of short tracks to climb the ranks on release.
He’s not alone - artists like Drake, Taylor Swift, and Ariana Grande have put out albums with twenty-plus songs to capture massive streaming numbers on release. Wallen has just taken it to the next level. While Spotify is a race to the bottom, Billboard is an arms race, a competition to see who can write forty tracks on an album and get half a billion streams in a week. What a world.
For those who can’t find success on TikTok or Spotify Discovery, there’s outright scamming:
Beatdapp, a Vancouver-based company that builds fraud detection software for labels, distributors and streaming services all over the world, estimates that around 10% of global streams are fraudulent; while some of this activity is caught, that could mean that over $1 billion in royalties ends up in the wrong pockets.
Streaming fraud! Of course. What does that mean?
“Streaming fraud” is a rangy term that can encompass a variety of behaviors. These include uploading an unauthorized remix of a viral single, boosting play counts by streaming music through a hacked Spotify account or creating a bot network to drive streams to a pop hit (for chart purposes) or a white noise recording (for royalties).
Indie artists may pad their numbers to increase their chances of signing with a label (or to pay rent). More insidious fraudsters might steal tracks and use bot networks to divert royalties into their own pockets. There are a dozen different incentives for fraud because, again, it is so difficult to make any money in the music business. The most catchy viral track on TikTok may net a writer a hundred bucks. The same track on Spotify might be worth thousands, depending on its staying power.
Streaming’s race to the bottom and our culture’s throwaway attitude towards music has both democratized music - we’ve got a historic glut of new songs thanks in part to social media and streaming - and made the experience of making it much, much worse for artists.
Chatbots
Last month, desperate to keep up with the AI arms race, Google released its Bard chatbot. The CEO has been on a whirlwind propaganda tour, making (false) claims its bot is spontaneously learning new languages and touting its incredible processing powers.
Should the company have seen the pushback and exposure of Bard’s many flaws coming? Its employees certainly did in the run-up to launch:
One worker’s conclusion: Bard was “a pathological liar,” according to screenshots of the internal discussion. Another called it “cringe-worthy.” One employee wrote that when they asked Bard suggestions for how to land a plane, it regularly gave advice that would lead to a crash; another said it gave answers on scuba diving “which would likely result in serious injury or death.”
Obviously, Google launched it anyway, despite staff objections. Some in tech have criticized Google’s glacial pace to launch something as simple as a chatbot when the company has - behind the scenes - been a leader in AI development. The difference between a public-facing chatbot and an army of engineers tinkering with black-box advertising algorithms or data modeling software is obvious: when an ad script messes up, little harm is done. When someone on Twitter screenshots your bot giving deadly scuba instructions, it’s a major PR issue.
Bard isn’t the only chatbot with a loose grasp on the truth, but thanks to the WaPo we know more about the dataset it’s trained on than its more secretive competitors. Ranking sites by the number of times their data appeared in the set surfaced some obvious primary sources:
The three biggest sites were patents.google.com No. 1, which contains text from patents issued around the world; wikipedia.org No. 2, the free online encyclopedia; and scribd.com No. 3, a subscription-only digital library.
And some…less obvious:
Also high on the list: b-ok.org No. 190, a notorious market for pirated e-books that has since been seized by the U.S. Justice Department. At least 27 other sites identified by the U.S. government as markets for piracy and counterfeits were present in the data set.
It begs the question - what liability does a company like Google have for scraping a bunch of illegal sites? What about personal data?
Two sites in the top 100, coloradovoters.info No. 40 and flvoters.com No. 73, had privately hosted copies of state voter registration databases. Though voter data is public, the models could use this personal information in unknown ways.
What about original content from creators?
Not far behind were kickstarter.com No. 25, which lets users crowdfund for creative projects, and further down the list, patreon.com No. 2,398, which helps creators collect monthly fees from subscribers for exclusive content.
[…]
The data set contained more than half a million personal blogs, representing 3.8 percent of categorized tokens. Publishing platform medium.com No. 46 was the fifth largest technology site and hosts tens of thousands of blogs under its domain. Our tally includes blogs written on platforms like WordPress, Tumblr, Blogspot and Live Journal.
Where are chatbots drawing from to create, say, sermons or religious poems?
Sites devoted to community made up about 5 percent of categorized content, with religion dominating that category. Among the top 20 religious sites, 14 were Christian, two were Jewish and one was Muslim, one was Mormon, one was Jehovah’s Witness, and one celebrated all religions.
Ah, well I’m sure that’s fine. Muslims and Jehovah’s Witnesses having equal standing in the chatbot’s eyes can only be a good thing. Surely the dataset will filter out extremist views:
The top Christian site, Grace to You (gty.org No. 164), belongs to Grace Community Church, an evangelical megachurch in California. Christianity Today recently reported that the church counseled women to “continue to submit” to abusive fathers and husbands and to avoid reporting them to authorities.
The highest ranked Jewish site was jewishworldreview.com No. 366, an online magazine for Orthodox Jews. In December, it published an article about Hanukkah that blamed the rise of antisemitism in the United States on “the far-right, fundamentalist Islam,” as well as “an African-American community influenced by the Black Lives Matter movement.”
Ah, well, nevertheless. The problem with creating any large language model (LLM) is that it’s large, and the LLMs the most powerful chatbots need are impossibly large, so large that OpenAI has spent years subjecting its underpaid African content moderators to terrible content to get its bot in a presentable state. When big companies like Google with have access to the largest index of the world’s Internet content rush to put out an LLM? The effects could be disastrous.
Google ignored its employees and abandoned its ethical guidelines to rush Bard to market because its shareholders and executives (some of its biggest individual shareholders! surely no conflict there) got worried they’d lose market share to competitors, a concept they’re largely unfamiliar with. Unfortunately, a monopolistic market position comes with great responsibility, and it seems Google is not taking its role seriously when it comes to chatbots.
Robotaxis
Late last year I wrote:
Like many things tech is trying to sell us these days, given our current capabilities and the size and scope of the problem, an unmanned robot taxi future remains science fiction, no matter how many billions we feed into the shredder.
Unfortunately for the citizens of San Francisco, California transportation regulators do not read this newsletter and seem hellbent on clogging their streets with robotaxis:
Driverless cars have completed thousands of journeys in San Francisco—taking people to work, to school, and to and from dates. They have also proven to be a glitchy nuisance, snarling traffic and creeping into hazardous terrain such as construction zones and downed power lines. Autonomous cars in San Francisco made 92 unplanned stops between May and December 2022—88 percent of them on streets with transit service, according to city transportation authorities, who collected the data from social media reports, 911 calls, and other sources, because companies aren’t required to report all the breakdowns.
92 unplanned stops doesn’t sound like too many, but the real number isn’t known because the companies aren’t required to report glitches. The WIRED article highlights a bigger problem with robotaxis causing transit delays - riders will blame bus and train systems rather than the tech companies behind the stoppages. Using San Francisco as a live beta testing facility is also cynically eroding public opinion of transit in an already horribly congested city that desperately needs better transit.
There is momentum behind the driverless car movement in SF, with its two main proponents having poured billions of (investor) dollars into the effort. Even if they were to succeed beyond all expectations and stop their cars from crashing, stopping, or getting in the way of transit, the end result is more cars on the road when more cars in a city with limited space isn’t going to improve conditions for residents. The article quotes an attorney suggesting that dedicated transit lanes would help solve the robotaxi problem but wouldn’t it also solve the need for a fleet of thousands of fucking robotaxis if people could simply hop a bus and get there faster?
Autonomous cars are still a solution in search of a problem. Autonomous tech for things like overnight freight deliveries or long haul trucking could be beneficial - improving lives and working conditions for truckers and helping reduce supply chain costs - but instead we’re spending billions automating more traffic jams.
WinAD
I often say that fraud is hard. It is hard because you have to concoct stories, fabricate lies and (sometimes) documents and not only weave those fictions, but keep track of it all. Some frauds go on for years, or decades, which just sounds exhausting. Living in a constant state of anxiety and fear, worried your fraud will be exposed and you’ll have to give back all your luxury cars or mansions or whatever.
Sometimes, successful businesspeople do frauds because they want to be more successful, and are jealous of their more affluent peers. Typically these frauds involve business and real estate deals, tax evasion, the kind of stuff you need money to perpetrate in the first place. Sometimes they are way more esoteric and complicated, bafflingly so.
Win AD is a company with a straightforward business model - they take public records data on NCAA Division I athletic teams and digitize it so other schools can search the data for competitive or research purposes. For example, if you are hiring a new employee at your university’s sports program, you might want to know what your competitors are paying. Or, you may want to know what another university spent on marketing last year. That sort of thing.
Two brothers named Ben and Drue Moore founded a company in 2009 to collate all this data, and by all accounts they have been wildly successful - something like 80% of all D-I universities use their tool which can cost up to $14,000 for a yearly subscription.
Not content to simply pocket their millions and live nice lives, the Moore brothers entered into a bunch of strange capers with an ex-felon who’s done time for bank fraud. They attempted to funnel Win AD’s profits through a dizzying number of strange schemes:
Their activities have been denounced in multiple courts, where opposing counsel and at least one judge have suggested the cousins are engaged in an illegal conspiracy. To date, Ben Moore has been sentenced to one 30-day jail sentence after being found in contempt for repeatedly ignoring discovery requests in the divorce proceedings with his now-ex-wife, where questions about the ownership and true value of Winthrop Intelligence have been central to the dispute.
Basically, the Moore brothers set up a bunch of corporations in Wyoming - a state friendly to business anonymity - and begin transferring ownership of their assets back and forth (and sometimes back to themselves) to hide them from the founder’s wives and the courts.
It gets really weird when the entities begin suing each other:
In essence, Ben Moore was suing Drue Moore, who had either ownership or controlling interest in each of the defendant companies. But since Ben had an interest in the companies through the Moore Family Office, he was also suing himself.
Moore companies accrued judgements against other Moore companies, which I guess they used to avoid taxes?
Handwritten notes and a diagram bearing [Scott] Brooks’ and [Ben] Moore’s penmanship detailed a plan whereby Winthrop Intelligence would pay a portion of a theoretical future court judgment to the Moore Family Office. The family office would then use that money to “pay off” a note payable obligation to Winthrop; loan Winthrop additional funds; and make a partial settlement payment to Brooks. “Now go around & around,” Brooks wrote, perhaps suggesting that this scheme was replicable.
Once their companies owed each other money via legal settlement, they’d funnel it offshore to avoid taxes and hide it from the wives, a central theme:
In a separate flow chart filled with colorful stickies, Brooks diagrammed how money could flow offshore from the Wyoming LLCs, to Davis’ law firm accounts, and then to a company in the United Kingdom.
Somewhere in the midst of all this were a hard cider company involving Scott Brooks’s son which they ran into bankruptcy (to secure more legal settlements) and a messy divorce that left a judge trying to unravel Ben Moore’s cloudy finances to levy fines and fees against him.
You can read the article and judge for yourself, but what jumps out to me is how complicated these two ostensibly successful dudes made their lives, all because they wanted to be more rich, or pay less taxes, or because a charismatic con man showed up in their lives and swore that if they just created twenty-seven LLCs in Wyoming and used a series of lawyers to funnel their money offshore by repeatedly suing themselves, it would all magically work out. For most people the hardest part is starting a successful business; for the Moore brothers, it was trying to orchestrate an elaborate fraud and keep their stories straight.
Short Cons
WaPo - “Zeidman had examined Lindell’s data and concluded that not only did it not prove voter fraud, it also had no connection to the 2020 election. He was the only expert who submitted a claim, arbitration records show.”
CNN - “The music group sent urgent letters in April to streaming platforms, including Spotify and Apple Music, asking them to block artificial intelligence platforms from training on the melodies and lyrics of their copywritten songs.”
Insider - “But while viewers can't get enough of the show, some cast members said being on "Love Is Blind" was a traumatic experience. They were so exhausted that they'd sometimes fall asleep during dates. Many were rarely spotted without an alcoholic beverage in hand.”
Insider - “The cost of room and board at many public universities is increasing faster than the rate of tuition, causing students to rack up deliriously high student loans and in some cases, leading them to drop out altogether.”
NBC News - “In New York City, a sample of 1,656 Uber Eats listings showed that more than 1 in 5 appear to be virtual, a trend repeated across each of the 12 cities NBC News examined, according to an analysis of restaurant listings.”
BuzzFeed News (RIP) - “In dozens of court documents reviewed by BuzzFeed News, cruise ship passengers say they have been dragged into cabins and raped, pushed into janitors’ closets and assaulted, and even attacked in the public corridors of ships. Likewise, parents and guardians have alleged that their children were molested by other passengers or crew members, plied with alcohol, and in some instances, abused by daycare staffers at onboard activity centers.”
Know someone planning on setting up an elaborate web of shell companies to hide money from the IRS and their spouse? Send them this newsletter!