Start Up No.2008: Google gets AI automating, OpenAI blocks astroturf effort, the trouble with media, Vietnam’s military trolls, and more

Poker isn’t a matter of luck. Which is how one writer had a mother who supported her family by playing for money when her husband died suddenly. CC-licensed photo by slgckgcslgckgc on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

A selection of 10 links for you. Think of some media. I’m @charlesarthur on Twitter. On Mastodon: Observations and links welcome.

Google to use new AI models for ads and to help YouTube creators • CNBC

Jennifer Elias:


Google’s effort to rapidly add new artificial intelligence technology into its core products is making its way into the advertising world, CNBC has learned.

The company has given the green light to plans for using generative AI, fueled by large language models (LLMs), to automate advertising and ad-supported consumer services, according to internal documents.

Last week, Google unveiled PaLM 2, its latest and most powerful LLM, trained on reams of text data that can come up with human-like responses to questions and commands. Certain groups within Google are now planning to use PaLM 2-powered tools to allow advertisers to generate their own media assets and to suggest videos for YouTube creators to make, documents show. .

Google has also been testing PaLM 2 for YouTube youth content for things like titles, and descriptions. For creators, the company has been using the technology to experiment with the idea of providing five video ideas based on topics that appear relevant.

With the AI chatbot craze speedily racing across the tech industry and capturing the fascination of Wall Street, Google and its peers, including Microsoft, Meta and Amazon, are rushing to embed their most sophisticated models in as many products as possible. The urgency has been particularly acute at Google since the public launch late last year of Microsoft-backed OpenAI’s ChatGPT raised concern that the future of internet search was suddenly up for grabs.


Well obviously Google would automate itself. A roundabout way of reducing headcount, perhaps? On the other side of the wall, advertising execs will of course be using ChatGPT and its siblings to churn out a zillion forms of the same message in text, picture and video form. Think of the 2016 Trump campaign which tried scores of slightly different Facebook ads and zeroed in on the ones that worked: like that, but more so.
unique link to this extract

OpenAI shut down DC company’s pitch to apply ChatGPT to politics • Semafor

Louise Matsakis:


Washington, D.C., company, FiscalNote, touted in a press release that it would use ChatGPT to help boost productivity in “the multi-billion dollar lobbying and advocacy industry” and “enhance political participation.”

Afterward, those lines disappeared from FiscalNote’s press release and were replaced by an editor’s note explaining ChatGPT could be used solely for “grassroots advocacy campaigns.”

A FiscalNote spokesperson told Semafor it never intended to violate OpenAI’s rules, and that it deleted that text from its press release to “ensure clarity.”

This is the first known instance of OpenAI policing how the use of its technology is advertised. The company last updated its policies in March, which now ban people from using its models for, among other things, building products for political campaigning or lobbying, payday lending, unproven dietary supplements, dating apps, and “high risk government decision-making,” such as “migration and asylum.”

OpenAI told Semafor that it uses a number of different methods to monitor and police when those policies are being violated. In the case of politics specifically, the company revealed it’s working on building a machine learning classifier that will flag when ChatGPT is asked to generate large volumes of text that appear related to electoral campaigns or lobbying.


A good catch, but you know there’s going to be more, done from the grassroots.

I deleted a big “KNOW MORE” from the text above which was encouraging you to read further, which you’d think the text would either achieve or not depending on how interesting it is, not whether you get told to “know more”. Speaking of which…
unique link to this extract

Incuriosity, Inc. • How Things Work

Hamilton Nolan:


The production of journalism in America depends to a remarkable degree on fooling rich people into thinking it’s a good idea to fund some publication, and then feverishly publishing as much stuff as possible before the rich person figures out that journalism is not a good investment.

An unfortunate consequence of this, though, is the profusion of publications designed from the ground up to appeal to the demographic of “business people who incorrectly imagine themselves to be ideas people.”

…The defining publication of the appeal-to-the-funder-type era is Axios, which did not invent the form I’m talking about, but which has certainly refined it to its highest/ lowest form. To make fun of the Axios house style is hardly original, but it is worth understanding it as the embodiment of a broken economic system in the journalism industry that relies purely on seducing people who don’t read:

Bold. This makes things look important whether they are or not.

• Bullet points are the baby food of language. Busy executives must have their brain food mashed up and put on tiny bite-sized spoons.
• Here’s another bite. Open up!

On the other hand: The false equivalency of “both-sidesism” has long been one of the plagues of journalism, causing good and bad ideas to be presented as equally meaningful, as a quick and dirty way to achieve the veneer of impartiality. Instead of eradicating this harmful tendency, imagine if you boiled it down into a thick, black concentrate, and then used that goo to write bold bullet points. Business people fucking love that shit. “Much to consider,” they nod.

Number. Statistics lend the appearance of validity to arguments regardless of their provenance.

• The American Enterprise Institute found that 63% of business guys aren’t maximizing their productivity.
• Much to consider.

Go deeper. Axios’s financial success in a hard media environment has attracted notice. Now there is also Semafor, which raised a ton of money to make what is in essence “Global cosmopolitan Axios.”


What’s hilarious about the Axios stuff is that you can delete the headings and •s and it reads just like a normal news story.
unique link to this extract

The Vietnamese military has a troll army and Facebook is its weapon • Rest of World

Danielle Keeton-Olen:


On Christmas Day 2017, Vietnam’s defense ministry announced a military group devoted to policing the country’s internet, called Force 47. In the five years since, pro-government trolls have been a persistent presence on the side of the regime, operating more or less freely across major platforms like Facebook and YouTube. As speech laws tighten in countries like India, Turkey, and Thailand — and platforms lose interest in pushing back — the trolls consistently and successfully harass activists and journalists posting on Vietnamese Facebook, providing a troubling model for how censorship can flourish within social media, even reaching beyond national borders.

In Vietnam, the fight has taken place largely outside the usual channels of law enforcement requests and court orders. In the first half of 2022, Facebook reported just under 1,000 takedowns based on local laws in Vietnam — above average, but still far fewer than neighbors like Taiwan, Thailand, and Indonesia. Over the same period, Facebook reported only one government request for user data within the country. By conventional measures, the Vietnamese government is not doing that much to restrict its citizens on Facebook.

But according to local opposition groups, activists and reporters, these numbers conceal a far more aggressive campaign of mass reporting of any groups that question or critique the government. Michel Tran Duc, the advocacy director for the pro-democracy group Viet Tan, told Rest of World he has to dispute a community standards violation on Facebook at least once a month. Michel is then forced to appeal the decision through Facebook — a slow and difficult process.


This is a story that’s familiar if you’ve read Social Warming: there’s coordinated effort and the distant moderators aren’t able to act soon enough or understand the problem. So bad state actors get away with it.
unique link to this extract

Introducing The Messenger Scale • The Messenger


Introducing The Messenger Scale, a new system designed to cut through the noise and help you understand what really matters in the news.

It’ll be like the “Richter scale” for measuring earthquakes, but in this case we will be assigning a simple 1-10 number based on input from our panel of more than 80 “news seismologists” from the worlds of politics, policy, law, history, academia and media. Our panel spans the entire political spectrum in order to provide readers with a balanced response to major news events.

…How does it work? When a news event happens, we’re hitting up our group with a simple question: On a scale of 1-10, how much do you think this event matters?

We purposefully kept the question vague and urged everyone to interpret it based on areas of expertise, understanding of the news cycle and where this particular event fits into the wider arc of history. 

The higher the number on The Messenger Scale, the more our panel thinks the event matters. The lower the number, the less they think the event matters.

We are averaging the responses to generate a number score down to the tenth decimal. We are keeping individual number ratings anonymous to encourage complete honesty, though who’s participating in The Messenger Scale will be public.


Since you’re wondering, 10 is big and/or bad. The January 6 insurrection was a 9.8, apparently. (Is a 10 a presidential assassination? How about aliens landing or a meteor strike – an 11, logically?)

The Messenger is a new website that has been described as “what if an entire website was a chumbox” (the junk links you see in a box below the story on a downmarket site). I liked the comment by Rusty Foster at Today In Tabs: “every new publication is looking for a moat, but only The Messenger had the guts to start by drowning in theirs.”
unique link to this extract

Fighting misinformation: using our trusted voice to good effect • Science Media Centre

Fiona Fox is the PR manager of the Centre:


Scientists at bodies that are wholly owned by, or arm’s-length from, a government department are not free to speak to journalists. There may be some managed media interventions occasionally with nominated spokespeople, but if any of any of you are scientists working for the NIBSC, the MHRA, UKHSA, the FSA, APHA, RCE, and many more like these, you will certainly not be encouraged to answer the phone to journalists, or engage regularly in ongoing controversies. And unfortunately, you will not be allowed to join the SMC’s database of experts providing us with comments on breaking news or new research.

And unlike my previous description of the changing culture, things are not getting steadily better in these arenas. Indeed, more and more science is being drawn into this heavily controlled and risk-averse culture. University or research institutes who are commissioned to do pieces of science for government often now have contracts which give political press officers leadership of the communications. University academics who have academic freedom are often subject to constraints if they agree to sit on a scientific advisory group to government. Both the former Chair and current CEO of UKRI have testified to excessive controlling tendencies from government departments, and we now rarely see the executive chairs of the Research Councils speaking openly in the news; a development that deeply saddened the late Professor Sir Colin Blakemore, who had often used his voice as head of the MRC to great effect.

When I raise this issue with scientists close to government, they invariably agree with me that we are missing out on some great scientists, but caution that the culture is entrenched and would simply be too hard to change. I’m sure that is true. But I think before we concede defeat, we should first ask ourselves collectively if it’s worth trying. As I have described, we have already collectively changed the culture of academic science to great effect. If we think misinformation is a serious barrier to progress in societal discussions, then surely we need to look hard at how we liberate more scientists to join the fight against it.


Those initials in full: National Institute for Biological Standards and Control, Medicines and Healthcare products Regulatory Agency, UK Health Services Agency, Food Standards Agency, Animal and Plant Health Agency, Radiation and Chemical and Environmental Hazards directorate. Colin Blakemore was an excellent spokesman for science.
unique link to this extract

How Big Oil is manipulating the way you think about climate change •

Kathleen Dean Moore:


Thirty-eight rail cars filled with vinyl chloride derailed and caught fire in East Palestine, Ohio. Vinyl chloride, a flammable petroleum product, is a potent carcinogen. When it is burned, it creates dioxin, another nasty carcinogen that now permeates the town. A familiar pattern followed: lamentations over the derailing; a cascade of reporters; a debate in Congress. Finally, politicians, commentators and outraged citizens all posed these questions: how will we punish the railroads? And how can we make railroads safer?

Those are the wrong questions. What I want to know is why would any sensible people allow the US petrochemical industry annually to produce 7.2 million metric tons of a poison that causes liver, lung, and brain cancer, and to distribute it as polyvinyl chloride in water pipes, gutters, rubber duckies, and My Little Pony dolls?

Another surprising example: In an effort to reduce the town’s use of fossil fuels, the city of Eugene, Oregon prohibited natural gas infrastructure in new residential construction. These types of prohibitions prompted a similar brand of handwringing — the question being posed in op-eds and comments sections running along the lines of, “How can anyone ask us to sacrifice our gas stoves, just to cut carbon emissions?”

That’s the wrong question. What I want to know is what sacrifices we are already making to support a fossil-fuel industry that earned $4 trillion in global profits last year, an industry whose control over us extends even to how we cook bacon-and-eggs. As ecologist Carl Safina said: “We are sacrificing our money, sacrificing what is big and permanent, to prolong what is small, temporary, and harmful. We’re sacrificing animals, peace, and children to retain wastefulness – while enriching those who disdain us.”


Questions worth asking. A worthwhile read too for the derivation of the phrase “red herring”.
unique link to this extract

Google, Facebook would pay “journalism fee” under California Bill • Bloomberg Law

Titus Wu:


California lawmakers are trying to advance legislation that would require digital platforms like Google and Facebook to pay a share of advertising revenue to media outlets after similar federal efforts have stalled.

The state legislation (A.B. 866) mirrors a federal measure in its attempt to help financially rescue local journalism companies and organizations. Newspapers rely heavily on social media and search engines to drive digital traffic as traditional print advertising revenues have disappeared. Meanwhile, titans like Google and Meta Platforms Inc. have controlled up to more than half of the online advertising market in recent years.

“These dominant digital ad companies are enriching their own platforms with local news content without adequately compensating the originators,” said Assemblymember Buffy Wicks (D) when announcing her bill in March. “It’s time they start paying market value for the journalism they are aggregating at no cost from local media.”

Big Tech, however, is characterizing the payments as a “link tax” that would disrupt the free and open nature of the internet.

The legislation faces strong opposition from a coalition of tech firms and even some journalism organisations, which contend the bill would actually harm local media and raises First Amendment issues. Backers of the measure also are wrestling with the logistics of imposing a proposed fee for original content as the bill awaits an Assembly floor vote.

Media groups in support of the bill say such payments would be an important financial lifeline for them to help them cover local issues from city council meetings to breaking news. Over 100 California newspapers have shuttered in the last decade as big tech companies have become the gatekeepers of how readers access news, said Emily Charrier, chairperson for the California News Publishers Association. She also serves as editor and publisher of the Sonoma Index-Tribune and publisher of the Petaluma Argus-Courier.


unique link to this extract

Tipping at self-checkout has customers crying ‘emotional blackmail’ • WSJ

Rachel Wolfe:


Zero interaction with employees during a transaction no longer guarantees freedom from the moral quandary of how much to tip.

Prompts to leave 20% at self-checkout machines at airports, stadiums, cookie shops and cafes across the country are rankling consumers already inundated by the proliferation of tip screens. Business owners say the automated cues can significantly increase gratuities and boost staff pay. But the unmanned prompts are leading more customers to question what, exactly, the tips are for.   

“They’re cutting labor costs by doing self-checkout. So what’s the point of asking for a tip? And where is it going?” says Ishita Jamar, a senior at American University in Washington, D.C., who has noticed more self-serve tip cues at restaurants she frequents.

Tipping researchers and labor advocates say so-called tip creep is a way for employers to put the onus for employee pay onto consumers, rather than raising wages themselves. Companies say tips are an optional thanks for a job well done.

Businesses “are taking advantage of an opportunity,” says William Michael Lynn, who studies consumer behavior and tip culture as a professor at Cornell University’s Nolan School of Hotel Administration. “Who wouldn’t want to get extra money at very little cost if you could?” 


Certainly not American companies, which avoid paying proper wages by expecting customers to make up the difference between what staff are paid, and should be paid.
unique link to this extract

My mother the poker shark • Esquire

Ian Frisch:


My mother had first started playing poker for the fun and for the intellectual challenge. Returning to competition twenty years later, she rediscovered old pleasures. She was playing not only to make money but also as an emotional escape. At the table, she wasn’t a single mother without a steady job mourning her husband’s death. It was the only place she felt comfortable playing the villain, cutthroat and cruel, lying to strangers’ faces and getting paid for it. “I love having a nemesis at the table,” she once told me. “It gives me purpose.” To this day, at every table, she picks a player and slowly, steadily, hand by hand, tries to destroy them.

To some people, poker is just a card game, a way to pass the time. For me and my mother, it’s a window into our identity, our way of understanding a world that at times can seem unforgiving. I began joining my mother in basement games around town in 2003, when I was sixteen. Ever since, poker has formed a bond between us, a mutual love, a prism through which I can see her not just as my mother but as a three-dimensional person who carries deep heartache and immense responsibility. Though it took me years to realize it, I now understand exactly how high the stakes were each time she sat down at a card table: It was the only way she knew how to keep living.


A lovely little vignette of a mother and son united by grinding other people down. Fairly.
unique link to this extract

• Why do social networks drive us a little mad?
• Why does angry content seem to dominate what we see?
• How much of a role do algorithms play in affecting what we see and do online?
• What can we do about it?
• Did Facebook have any inkling of what was coming in Myanmar in 2016?

Read Social Warming, my latest book, and find answers – and more.

Errata, corrigenda and ai no corrida: none notified

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.