…mornin’ all…& welcome to today’s installment of “do we have to?”

Over the past decade, the focus on AI in Chinese society and digital culture has grown. Since the Covid-19 outbreak, AI implementations in schools, office buildings and factories have rolled out in fast forward.

AI facial recognition is employed in everything from public security to payment technology; smart glasses and helmets make it easier for many workers to perform their tasks; and intelligent robots have become a common sight in China’s service industry, in malls, restaurants, and banks.

There seemed little doubt over who would win the tech race between the eagle and the dragon; but then came ChatGPT.

It took months for China to launch its own alternative, models that seemed to lag behind their western variants in multiple ways. Even the minister of science and technology acknowledged that China’s chatbots were struggling against their US competition and Chinese internet users were left asking why – given that China was meant to dominate the AI era.

Experts and bloggers proposed different answers: some suggested China was not the first to launch a ChatGPT-like product because tech startups in China tend to focus on fast applications rather than lengthy research and development. Others said that language model training in China was harder due to the rich and complex nature of the language.

But many seemed to agree that it was the political sensitivities and the Chinese online environment – which is closely monitored and subjected to censorship – making development of ChatGPT-like platforms more challenging in China.

In the summer of 2023, Chinese authorities proposed rules for generative AI, mandating that AI-generated content, whether images or text, must align with the “core values of socialism” and must not undermine state authority, harm national unity or spread false information. AI service providers were also mandated to prevent users from becoming too dependent on their services.

https://www.theguardian.com/commentisfree/2023/apr/22/can-china-keep-generative-ai-under-its-control-well-it-contained-the-internet

But does all of this mean that China’s AI revolution is losing momentum? Not at all.
[…]
Thanks to new AI technology, business owners can now purchase their own deepfake influencers to work and sell for them at all hours of the day, offering unprecedented opportunities to small Chinese entrepreneurs. The popularity of these deepfakes exploded on e-commerce streaming platforms in 2023.

New digital employees can not only answer customers’ questions but also gauge if they are smiling, knowing when to keep their answers short. One such digital avatar was named employee of the year at a Chinese real estate developer.
[…]
Meanwhile, Chinese authorities are working with big tech companies to make Communist party messages as appealing and accessible as possible to Chinese people of all ages. State newspaper the People’s Daily has introduced a virtual presenter.

https://www.theguardian.com/world/2024/jan/09/in-the-race-for-ai-supremacy-china-and-the-us-are-travelling-on-entirely-different-tracks

…&…call me a block-quoting hypocrite…but…I have a sneaking suspicion that the not-making-scads-of-cash-off-it aspect of my approach to fair use probably has a lot to do with not getting sued by the NYT…the not being a world famous brand name trying to be the google of AI probably factors in there too…in a security-through-obscurity sort of a way

The developer OpenAI has said it would be impossible to create tools like its groundbreaking chatbot ChatGPT without access to copyrighted material, as pressure grows on artificial intelligence firms over the content used to train their products.
[…]
Last month, the New York Times sued OpenAI and Microsoft, which is a leading investor in OpenAI and uses its tools in its products, accusing them of “unlawful use” of its work to create their products.

In a submission to the House of Lords communications and digital select committee, OpenAI said it could not train large language models such as its GPT-4 model – the technology behind ChatGPT – without access to copyrighted work.

“Because copyright today covers virtually every sort of human expression – including blogposts, photographs, forum posts, scraps of software code, and government documents – it would be impossible to train today’s leading AI models without using copyrighted materials,” said OpenAI in its submission, first reported by the Telegraph.

…I’ll skip past the part where that pronouncement seems to be almost wholly dependent on your definition of terms…but…& you don’t have to have read nearly as many sci fi books of potentially dubious quality as I have to know that one of the mainstays of the sort of AI we don’t have but a lot of people like to think would be pretty neat to have about the place…along with their hoverboards, lightsabers & flying cars…is that once it has access to “the internet” we assume it’s working from a database of the entire sum of human knowledge & has the ability to understand all of it as well as anyone up to & including the most expert mind available…which is right handy if you’re an author looking to inject a little exposition or in need of a deus in machina…or whatever…& that makes it easy for people to follow the general line of reasoning that says “if you tell us no then you never get *the thing*”…but…in a dismally prosaic fashion…the whole business seems…in context…to be a negotiation about the price point for the licensing arrangements rather than the grand principles or the speculative fiction goalposts getting invoked

It added that limiting training materials to out-of-copyright books and drawings would produce inadequate AI systems: “Limiting training data to public domain books and drawings created more than a century ago might yield an interesting experiment, but would not provide AI systems that meet the needs of today’s citizens.”

…they’re…not necessarily wrong, as it goes…vanishingly few of us can, when you consider how long it takes to become an expert in any of the myriad ways that people are with regard to this or that, claim to have really done the work…we rely on one of the “value-added” aspects of a whole host of stuff being that the homework did get done by the folks that produced it…& if you think of the less artificial version of chat…it’s always possible the conversation includes someone who did buy & read the NYT…& whose mind can serve up responses that include that in the datasets from which they are…organically…derived…& we wouldn’t think it made sense for them to wind up owing the NYT some sort of commission-style micro-payment as a result…&…if we didn’t have search engines the internet would be even less navigable than it is…so…if we want AI to be able to include being a natural language interface that can do the tech-y lifting for us…it’s not an entirely specious argument to admit that the economies of scale involved do that “approaching but never quite reaching infinity” thing I’m pretty sure mathematicians have a name for…that I’m too human & lazy to bother looking up right now…& very possibly make it impossible to not go bust implementing before you get to the promised land that makes it all worth while…you know…barring accidents

Elsewhere in its House of Lords submission, in response to a question about AI safety, OpenAI said it supported independent analysis of its security measures. The submission said it backed “red-teaming” of AI systems, where third-party researchers test the safety of a product by emulating the behaviour of rogue actors.

OpenAI is among the companies that have agreed to work with governments on safety testing their most powerful models before and after their deployment, after an agreement struck at a global safety summit in the UK last year.

https://www.theguardian.com/technology/2024/jan/08/ai-tools-chatgpt-copyrighted-material-openai

…but…whether you think it’s all window dressing for the operative bottom line…or the end is some sort of quasi-religious reward for blind faith in the power of progress that will produce a machine capable of saving humanity & its autonomous cosmic ark from…well…us lot…or you trust WaPo’s shortlist that ran the other day of the half a dozen cases lining up to be the textbook examples of how we call the balls & strikes on the former in keeping with our thoughts on the latter…while praying the line we’re taking follows aristotle’s golden mean & isn’t in fact merely a yellow-brick road

[…not for nothing but a couple are federal cases against google for antitrust stuff…but…do we want search results based on not having looked everywhere…& if not…could we maybe…I dunno…look into whether or not it’s helpful to have the results of the same search made by different people…or, hell, the same damn person on two different devices…or two different browsers…or two different search engines…be entirely different…mostly I tend to think in some ways that’s a good thing & in others it’s not…but that it wouldn’t really be a problem if people understood the tools well enough to be able to fit them to the tasks they’re good at…&…historically…humans aren’t exactly built that way for the most part]

…or…maybe a primrose path

A.I. is set to advance at a rapid rate, becoming more powerful and spreading into the physical world. [NYT]

In the hands of anonymous internet users, A.I. tools can create waves of harassing and racist material. It’s already happening on the anonymous message board 4chan. [NYT]

Judge in Trump election subversion case targeted by fake 911 ‘swatting’ call [Guardian]

‘Designed to terrorize’: Maine official who removed Trump from ballot describes recent threats [Guardian]

…uh huh…cool, cool, cool…it’s the principle of the thing, you say?

As the mother of an 8-year-old, and as someone who’s spent the past year experimenting with generative A.I., I’ve thought a lot about the connection between interacting with one and with the other. I’m not alone in this. A paper published in August in the journal Nature Human Behaviour explained how, during its early stages, an artificial intelligence model will try lots of things randomly, narrowing its focus and getting more conservative in its choices as it gets more sophisticated. Kind of like what a child does. “A.I. programs do best if they start out like weird kids,” writes Alison Gopnik, a developmental psychologist.

I am less struck, however, by how these tools acquire facts than by how they learn to react to new situations. It is common to describe A.I. as being “in its infancy,” but I think that’s not quite right. A.I. is in the phase when kids live like tiny energetic monsters, before they’ve learned to be thoughtful about the world and responsible for others. That’s why I’ve come to feel that A.I. needs to be socialized the way young children are — trained not to be a jerk, to adhere to ethical standards, to recognize and excise racial and gender biases. It needs, in short, to be parented.
[…]
It’s not enough to simply tell children what the output should be. You have to create a system of guidelines — an algorithm — that allows them to arrive at the correct outputs when faced with different inputs, too. The parentally programmed algorithm I remember best from my own childhood is “do unto others as you would have done unto you.” It teaches kids how, in a range of specific circumstances (query: I have some embarrassing information about the class bully; should I immediately disseminate it to all of my other classmates?), they can deduce the desirable outcome (output: no, because I am an unusually empathetic first grader who would not want another kid to do that to me). Turning that moral code into action, of course, is a separate matter.

Trying to imbue actual code with something that looks like moral code is in some ways simpler and in other ways more challenging. A.I.s are not sentient (though some say they are), which means that no matter how they might appear to act, they can’t actually become greedy, fall prey to bad influences or seek to inflict on others the trauma they have suffered. They do not experience emotion, which can reinforce both good and bad behavior. But just as I learned the Golden Rule because my parents’ morality was heavily shaped by the Bible and the Southern Baptist culture we lived in, the simulated morality of an A.I. depends on the data sets it is trained on, which reflect the values of the cultures the data is derived from, the manner in which it’s trained and the people who design it. This can cut both ways. As the psychologist Paul Bloom wrote in The New Yorker, “It’s possible to view human values as part of the problem, not the solution.”

…we sometimes like to call the less consciously determined operations of our brains instinctive…& everyone knows any detective worth their salt follows their gut…but…we also call that strata of our mental geology “sub-conscious”…which…seems like a term you could apply to a dolled-up mechanical turk we’ve elected to dub “intelligent”…not that I imagine anything about any of that would be significant or anything…non-artificial intelligences routinely mistake tone or misinterpret nuance, anyway…so what does it matter if there’s already a surplus of purportedly autonomous things laying claim to intelligence they neither model nor display…I’m sure it’ll be fine…dumbass assholes who aren’t nearly as clever as they think they are haven’t caused us any major upsets thus far in human…sorry…*checks notes*…uhhhhh…never mind…moving swiftly on

A similar dynamic emerges when A.I.s that have not been designed to tell only the truth calculate that lying is the best way to fulfill a task. Learning to lie as a means to an end is a normal developmental milestone that children usually reach by age 4. (Mine learned to lie much earlier than that, which I took to mean he is a genius.) That said, when my kid lies, it’s usually about something like doing 30 minutes of reading homework in four and a half minutes. I don’t worry about broader global implications. When A.I.s do it, on the other hand, the stakes can be high — so much so that experts have recommended new regulatory frameworks to assess these risks. Thanks to another journal paper on the topic, the term “bot-or-not law” is now a useful part of my lexicon.
[…]
I’m not an A.I. pessimist generally. My p(doom) estimate — the probability that A.I.s will be the end of us — is relatively low. Five percent maybe. Eight on days when an A.I.-powered autocorrection tool inserts appalling typos into my work. I believe A.I. can relieve humans of a lot of tedious things we can’t or don’t want to do, and can enhance technologies we need to solve hard problems. And I know that the more accessible large language model applications become, the more possible it will be to enable them to parse moral dilemmas. The tech will become more mature, in both senses.

But for now, it still needs adult supervision, and whether the adults in the room are equipped to do that is up for debate. Just look at how viciously we fight over how to socialize real children — whether, for example, access to a wide array of library books is good or bad. The real danger is not that A.I.s become sentient and destroy us all; it’s that we may not be equipped to parent them because we’re not mature enough ourselves.

I Finally Figured Out Who ChatGPT Reminds Me Of [NYT]

…gotta love people…we just can’t seem to help ourselves…we talk about one thing & then argue that means we did or didn’t say this or that about an entirely different thing the shape of which is easier to handle to our way of thinking…& self-appointed “authorities” feel the need to tell everyone who’ll listen that not only can a thing only mean one thing but they’re here to tell us all what that one thing is…just in case…I dunno…we might fail to see things their way & inadvertently stumble into understanding something about the world that eludes their blinkered view of what they consider to be a didactic reality…that…just like literally everybody else’s…only exists the way they think it does within the confines of their mind…but…if we start carving up the infinite distinctions between the intersubjective & the objective…we’ll be here all day…& I haven’t got all day…so…much as I’d consider it a legitimate recreational pursuit…I’d best be in search of the point I assume some bit of my mind was presumably driving at when it started typing this morning

These are the men that try The Times’s soul.

With the disreputable Donald Trump challenging the disfavored President Biden, the 2024 race has become the embodiment of Oscar Wilde’s witticism about fox hunting: “the unspeakable in pursuit of the inedible.”

Bleeding young and nonwhite voters, the president finally heeded Democrats urging him to “get out there,” as Nancy Pelosi put it, and throw some haymakers at Trump.

Biden flew to Pennsylvania on Friday to visit Valley Forge and make a pugnacious speech invoking an earlier moment when we were fighting against despotism and clinging to a dream of a democracy.

In a discontented winter during the American Revolution, George Washington tried to inspire his downtrodden troops at Valley Forge by having Thomas Paine’s “The American Crisis” read to them.

“These are the times that try men’s souls,” Paine wrote, adding, “Tyranny, like hell, is not easily conquered.”
[…]
If we bow down to a wannabe dictator who loves dictators, who echoes the language of Nazi Germany, who egged on the mob on Jan. 6 and then rewrote the facts to “steal history” just as he tried to steal the election, Biden wondered, what does that say about who we are?

Rejecting Trump’s campaign of grievance, vengeance, malignance and connivance, the president said, “We never bow. We never bend. We speak of possibilities — not carnage. We’re not weighed down by grievances. We don’t foster fear. We don’t walk around as victims.”

https://www.nytimes.com/2024/01/06/opinion/trump-biden-election.html

…ummm…*checks notes*…if that were a ChatGPT-generated response I’d be tempted to call it a hallucination…I mean…I’m not blind to at least many of the ways in which you might arrive at crafting it as a statement designed to have the desired effect on the listeners you’re hoping to target it towards…& a bunch of those include a host of internalized examples of wishful thinking & delusional but fond beliefs in things like exceptionalism or the power of the ruggedly individualistic approach to cherry-picking from the menu vis à vis who has to hold up which end of what social-contract-wise…I’d even expect that as a baseline even the average person would in fact be able to instinctively parse that stuff to the point of being able to claim they “get” what the man was saying…but…& I can’t help getting a tad hung up on this part…even if enough of those minds were to bounce that off one another in a sufficient profusion of ways in a sufficient profusion of places to mean their eavesdropping artificial counterpart can brute force a simulacrum of engaging in the same exercise via…effectively…the pattern-recognition version of a brute-force attempt at code-breaking…well…it keeps tripping me up…call it a signal:noise issue…or degradation…think of a moderately intelligent mind as a master copy…as that understanding becomes further & further removed from the mind that did the original understanding to be passed off as the understanding of others…you get…fidelity issues…interpolated artifacts that weren’t part of the original…& an inverted ratio between claims of certainty & the underlying reality…but then, I’m implicitly biased…what with thinking it’s almost always more interesting to ask someone why they think something than it is to tell them you know better & will brook no dissent…& I’m comfortable admitting I haven’t made my mind up about most things in particular so much as come to a few general conclusions about matters in the abstract which don’t always seem like a good fit for the reality I find myself immersed in…which has given me no end of trouble over the years with the way that…predominantly if not exclusively…the mode of expression in the US is to routinely present as axiomatic things that do not so much beg the question as beggar the poor bugger…not to mention bugger the beggar…if you’re feeling nasty…or merely solipsistic…I mean…oh, what’s the point…I’m out of time as it is…& I actually do have things I have to get done…places to be, people to see…all that fun stuff…but…by way of an example of the kind of thing I’m talking about

“This was the least defensible op-ed I can remember ever seeing the NYT run, made all the worst by the fact that it was written by a staffer, who specializes in these speculations,” Chris Willman, the chief music critic at Variety, wrote on Twitter. (In 2022, Marks wrote a guest op-ed essay for the Times speculating on Harry Styles’s sexuality, as well.)

Wilman’s tweet was reposted by Chely Wright, a queer country singer whose struggles to publicly come out in her career were mentioned in Marks’s piece. “I think it was awful of [the New York Times] to publish,” she wrote. “Triggering for me to read – not because the writer mentioned my nearly ending my life – but seeing a public person’s sexuality being discussed is upsetting.”

https://www.theguardian.com/music/2024/jan/08/taylor-swift-nyt-opinion-sexuality

…now…I don’t have anything invested…in any sense of the term…in the putative sexuality of that (or any other) artist beyond the possibility it adds a potential dimension to the context in which to view their work & to try to understand…to borrow an analogy…the intent behind the hand that made the particular watch we’ve stumbled across while trekking the cultural desert…but…on the one hand there’s no actual need for the artist to bring that context when the beholder can go ahead & do it all by themselves…consciously or unconsciously…entirely independently…in ways that it helps to know the academic definition of “a necessary connection” to really get to grips with…which is a whole other nested set of gearing & leverage we mostly feel fine using to get us from A to B…or even alpha to omega…while treating the same way we do all the other stuff under the hood that gets us there…which…again…I lack the time to get properly into the weeds about…but there were two sets of reasons I suggested I thought the NYT piece was worth a read & they only tangentially require there to be any attention paid to the putative sexual preferences of a particular slim white pop-princess…or…in another time & place…thin white duke

…one was that I found it well-written in the sense that it read well…another was that it provided ample evidence of the author having something beyond a well-grounded knowledge of the material…& yet another was that it did not purport to providing certainty about the uncertainty at the root of what was being considered but did do a better than average job of navigating with respect the different ways that different people for different reasons might be inclined to feel the need for that answer to be a particular one…while also saying…in more or less so many words…that there is some there, there…should you choose to see it…& that in the context of this particular individual…that seems highly unlikely to be the case unintentionally

…& some people…who aren’t me…would very possibly have trenchant opinions on whether or not it would be okay for all the stuff it references to exist within her work if she in fact is just some boring old cis-het exemplar of white privilege…but…personally…more so than if we were talking about bowie…or eminem…in a female artist…who came up through the nashville machine…for there to be a network of footholds across a spectrum that requires you to view their whole existence as a holistic exercise in performance intended as dialogue…one that not only allows such an argument to gain traction but actually supports more weight of scrutiny than many things people unhesitatingly declare themselves certain of left, right & center?

…if that’s not virtuosity…what is?

…try pulling that out of even the most unrestrained of LLMs the way people pull those opinions out of their ass on the daily…go on…there’s too few things that make me laugh…& every little helps