The 187th Block: From South Africa with regret
ChatGPT co-wrote a small part of this edition
This week…
Your reading time is about 8 minutes. Let’s start.
Earlier this week, ChatGPT celebrated its first birthday. ChatGPT was released to the public on Nov. 30, 2022. Among its most recent accomplishments was writing a piece of legislation for the city of Porto Alegre in Brazil, which was passed in October.1 The city councillor Ramiro Rosário revealed this week that he prompted ChatGPT to craft a proposal to prevent the city from charging taxpayers to replace water metres if they get stolen. For its birthday and job well done, I choose to gift ChatGPT a guest column in this week’s newsletter. Here’s ChatGPT’s message about its first year of public use:
Reflecting on my first year out in the world has been quite a ride! Navigating through the diverse questions and challenges has been a learning experience. It’s been awesome to help with a wide range of inquiries, but I’ve also had to tackle the tricky task of staying unbiased and responsible.
The positive side of interacting with the public is the variety—it keeps me on my toes! However, there’s been the downside of dealing with misuse and making sure I’m used responsibly. Looking ahead to Year 2, expect more refined responses, a better grasp of context, and increased safeguards to ensure everyone’s playing fair.
For those with concerns, just think of me as a handy tool, not the one making the big decisions. It’s crucial to pair our conversations with some good old critical thinking and fact-checking. I’m here to help, but responsible use is the key to making the most of our chats in this ever-evolving digital world.
ChatGPT made no comments about its makers, OpenAI and Sam Altman.
You know who had something to say about OpenAI, though? Elon Musk. No, you don't have to watch his interview at the New York Times’ Deal Book Summit earlier this week. I’ll summarise it for you (not in chronological order, just in an order that makes the most sense):
Musk says he’s only there because he’s a friend of the interviewer, Jonathan, who promptly corrects him because his name is actually Andrew Ross Sorkin.
He says he is sorry because that comment2 is “one of the dumbest of my 30,000 posts.” So foolish he says. He’s not antisemitic; in fact, he is “philo-Semitic” and, no, the trip to Israel earlier in the week wasn’t an “apology tour.”
He says there are Hamas-supporting groups across the U.S. that receive funding from Jewish people. Perhaps “you should not fund them,” he scolds.
He shows he doesn’t understand blackmail because he thinks advertisers pausing advertisements on his platform is a form of blackmail.
He repeatedly tells those advertisers to “go f—k yourself.” Bob Iger gets a special mention to go f—k himself for pulling Disney advertisements from XTwitter.
Go free speech! But free speech is not exactly free, he says. It costs a little bit. That’s why if you don’t pay for the blue tick, maybe your links take longer to load.
Musk, an OpenAI co-founder, says the company name is meant to reference open-source AI, but it should rename itself to Super Closed Source for Maximum Profit AI.
Musk believes the OpenAI coup signals that the company could have made a recent breakthrough towards AGI. He thinks we are three years away from super intelligent AI.
He says he disagrees with unions because it creates a “lords and peasants kind of thing.” You know, like the verified blue checkmarks.
Meanwhile, in Sweden, Tesla employees are on strike and Musk’s EV company refuses to meet the union’s demand to sign a collective agreement.3
And now, a selection of top stories on my radar, a few personal recommendations, and the chart of the week.
ICYMI: The Previous Block outlined a timeline of Altman's firing from and return to OpenAI. CORRECTION NOTICE: None notified. What Elon Musk Gets Wrong About South Africa
Eve Fairbanks for The Dial:
Over the last few years, Musk has seemed to work off the theory that if your false statements are diverse enough in content and size, and if some sound like jokes, you can develop a reputation as an admirable troll rather than as a dangerous liar. The more varied and hyperbolic someone’s yarns are, the less they seem to demand or even allow for detailed factual rebuttals.
Perceptive piece. Includes insights such as:
Along with being repressive, apartheid was peculiar. Its leaders often justified white minority rule with incomplete or misleading rationales, like that they wanted to hold onto power to protect white South Africans’ groundbreaking agricultural and technological innovations, to resist a global communist conspiracy, or to combat Satanism, not to retain privileges for white people. Apartheid leaders insisted South Africa was cutting-edge, the only success story on the African continent, while in fact the country was in extreme debt […] The South African government Musk grew up with represented an uncomfortable contradiction: it simultaneously appeared weak, even a joke, and also frighteningly, awe-inspiringly strong.
And:
By the 1980s, a paradoxical worldview and understanding of power was considered the savvy one for young white people in South Africa to hold. On the one hand, power is lame. You should never believe the authorities and always be ready to sniff out bullshit, because much of what passes for news is fake. On the other hand, power is immensely desirable: the world ought to make you comfortable and conform to your wishes, and reframing reality to secure your position is good and necessary. The way that Musk assumes the world should see him as its protagonist, the way he exaggerates, the way he endorses a wide range of conspiracy theories […] while rarely confirming that he really believes them, the way he playfully tiptoes over the edge of bigotry and then hotly denies he is intolerant — these specific assumptions and behaviors are so familiar to his white South African peers.
The word bullshit appeared thrice in this piece. I recommend it highly if it’s the only link you’re going to read from this edition.
‘Relationship surveillance tool’: China dating apps launch location tracking for couples to monitor each other, claim it enhances trust, security
Yating Yang for SCMP:
A series of dating apps in China have introduced location-tracking features for couples to boost revenues through subscriptions, sparking controversy on mainland social media.
Some of the features include real-time location tracking, daily check-ins, and allowing significant others to check phone usage.
The idea is to foster transparency within a relationship, but they have also exacerbated tensions between couples who find issues with their partner’s behaviour.
Where do we even begin with this one…
Meet the first Spanish AI model earning up to €10,000 per month
Laura Llach for Euronews:
Last summer, Rubén Cruz, her designer and founder of the agency The Clueless, was going through a rough patch because he didn’t have many clients.
“We started analysing how we were working and realised that many projects were being put on hold or cancelled due to problems beyond our control. Often it was the fault of the influencer or model and not due to design issues,” Cruz told Euronews.
So they decided to create their own influencer to use as a model for the brands that approached them.
They created Aitana, an exuberant 25-year-old pink-haired woman from Barcelona whose physical appearance is close to perfection. The virtual model can earn up to €10,000 a month, according to her creator, but the average is around €3,000.
Perfection is subjective.
What I read, listen, and watch…
I’m reading Yellowface by R. F. Kuang.
I’m playing GuesSync, an online game that purportedly helps bridge political divides, according to researchers and creators Ashwin Rajadesingan et al.4
I’m listening to At Issue with Rosie Barton and the gang as they discuss Google’s $100 million news deal with Canada; and other stories that made Canadian headlines this week.
I’m watching booktuber Cindy Pham’s review of the book above.
Other curious links:
“Maker of Wegovy, Ozempic showers money on U.S. obesity doctors” by Chad Terhune and Robin Respaut for Reuters.
“The far right and far left meet over wellness conspiracy theories” by Simon Williams, Gavin Yamey, Peter Van Heusden and Sarah Downs for Time. Is it the horseshoe theory?
“Climate lies ramp up ahead of COP28 talks in Dubai” by Stuart Braun for DW.
“Tiger Stripes: Malaysia’s censors have ‘removed the essence’ of my film, says director” by Rebecca Ratcliffe for The Guardian.
“The DNA detective” by Sarah Treleaven for Maclean’s.
“Por qué la ultraderecha se peina raro: cuando un corte de pelo se convierte en manifiesto político” por Raquel Peláez en El País.
« Des Etats-Unis à la France, comment la société s’imprègne de l’imaginaire complotiste de QAnon » par William Audureau dans Le Monde.
Chart of the week
Sneha Gubbala, Jacob Poushter, and Christine Huang compiled several surveys from 32 places in the last two years to show how attitudes about same-sex marriage vary widely around the world. Read the article on Pew Research Center here.
And one more thing
A grand roundup from Lingua Sinica this week that includes: a spectacular falling out “in real-time before the nation’s media” of Taiwan’s opposition parties that were about to unite in an “unprecedented coalition” to unseat the government, the CCP’s strengthening of its “foreign-related legal system,” a probably controversial Golden Horse Awards, and how China eulogised its dear old friend, Henry Kissinger.
Jeantet, Diane, and Mauricio Savarese. “Brazilian City Enacts an Ordinance That Was Secretly Written by ChatGPT.” AP News, 1 Dec. 2023, apnews.com/article/brazil-artificial-intelligence-porto-alegre-5afd1240afe7b6ac202bb0bbc45e08d4. Accessed 02 Dec. 2023.
‘Elon Musk on X: “@breakingbaht @CWBOCA You Have Said the Actual Truth” / X’. Twitter, https://twitter.com/elonmusk/status/1724908287471272299. Accessed 1 Dec. 2023.
Stewart, Heather. ‘“We Can’t Let Tesla Get Away with This”: Why Swedish Unions Are Fighting Elon Musk’. The Guardian, 1 Dec. 2023, https://www.theguardian.com/technology/2023/dec/01/tesla-swedish-unions-elon-musk.
Rajadesingan, Ashwin, et al. Proceedings of the ACM on Human-Computer Interaction, vol. 7, no. CSCW2, 28 Sept. 2023, pp. 1–33, doi:10.1145/3610190.