Start Up No.2032: US to restrict China cloud access, Twitter’s shonky paid API, CJEU prangs Facebook’s business model, and more


If things have gone as normal, sightings of UFOs will have peaked in the US on Independence Day. Guess why. CC-licensed photo by maxime raynal 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.


On Friday, there’s another post due at the Social Warming Substack at about 0845 UK time.


A selection of 9 links for you. Mysterious. I’m @charlesarthur on Twitter. On Mastodon: https://newsie.social/@charlesarthur. Observations and links welcome.


US looks to restrict China’s access to cloud computing to protect advanced technology • WSJ

Yuka Hayashi and John D. McKinnon:

»

The Biden administration is preparing to restrict Chinese companies’ access to US cloud computing services, according to people familiar with the situation, in a move that could further strain relations between the world’s economic superpowers.

The new rule, if adopted, would likely require US cloud service providers such as Amazon.com and Microsoft to seek US government permission before they provide cloud computing services that use advanced artificial intelligence chips to Chinese customers, the people said.

The Biden administration’s move on cloud services comes as China said Monday it would impose export restrictions on metals used in advanced chip manufacturing.

This high-stakes conflict over supply chain access to the world’s most advanced technology is escalating in the days ahead of Treasury Secretary Janet Yellen’s visit to China, a trip the Biden administration hopes will ease tensions. Yellen’s talks in Beijing are expected to touch on macroeconomic conditions in each country, as well as climate change and debt in the developing world.

The US’s proposed cloud restrictions are seen as a means to close a significant loophole. National security analysts have warned that Chinese AI companies might have bypassed the current export controls rules by using cloud services.

These services allow customers to gain powerful computing capabilities without purchasing advanced equipment—including chips—on the control list, such as the A100 chips by American technology company Nvidia.

“If any Chinese company wanted access to Nvidia A100, they could do that from any cloud service provider. That’s totally legal,” said Emily Weinstein, a research fellow at Georgetown Center for Security and Emerging Technology.

«

Oh, that is quite the loophole, isn’t it. Wonder how they’ll enforce it, though. Proxy hopping from one server to another, using shell companies.. it’s going to be quite the whack-a-mole game. Another little brick in the trade war wall.
unique link to this extract


Twitter’s API keeps breaking, even for developers paying $42,000 • Mashable

Matt Binder:

»

Twitter’s new API may now cost tens of thousands of dollars per month, but the service being provided to its customers appears to be worse than ever.

That’s the general sentiment among developers who are still part of the once-robust third-party Twitter app ecosystem. According to developers paying Twitter, since the switch over to Elon Musk’s paid API subscription plans, Twitter’s API has experienced frequent issues that make it extremely difficult to run their apps.

Twitter’s API issues have frustrated developers in each of Twitter’s new API access tiers. Those with Basic or Pro plans — paying $100 and $5000 a month for API access, respectively — have experienced unannounced changes to their plans, numerous bugs, and often receive zero customer support. And developers shelling out for Twitter’s Enterprise API Plan, which starts at $42,000 per month, are experiencing sudden outages and disappointing service considering the money they’re paying.

“Everything used to work fine before we started paying half a million per year,” shared one developer in a private Twitter developer group chat shared with Mashable.

«

As John Gruber points out, why – now you know this – would you ever pay for Twitter Blue (or, now, for Tweetdeck)? Amazing how Musk has transformed it from a reliable, low-earning must-use service into an unreliable, lossmaking service that people are keen to abandon.
unique link to this extract


CJEU ruling on Meta referral could close the chapter on surveillance capitalism • TechCrunch

Natasha Lomas:

»

A long-anticipated judgement handed down today by the Court of Justice of the European Union (CJEU) looks to have comprehensively crushed the social media giant’s ability to keep flouting EU privacy law by denying users a free choice over its tracking and profiling.

The ruling tracks back to a pioneering order by Germany’s antitrust watchdog, the Federal Cartel Office (FCO), which spent years investigating Facebook’s business — making the case that privacy harm should be treated as an exploitative competition abuse too.

In its February 2019 order, the FCO told Facebook (as Meta still was back then) to stop combining data on users across its own suite of social platforms without their consent. Meta sought to block the order in the German courts — eventually sparking the referral on Meta’s so-called “superprofiling” to the CJEU in March 2021.

Now we have the top court’s take and, well, it’s not going to spark any celebrations at Meta HQ, that’s for sure.

The CJEU has not only agreed competition authorities can factor data protection into their antitrust assessments (which sounds wonky but really is vital because joint-working rather than regulatory silos is the path to effective oversight of platform power) — but has signalled that consent is the only appropriate legal basis for the tracking-and-profiling-driven ‘personalized’ content and behavioral advertising that Meta monetizes.

«

It’s deeply complicated. Here’s the CJEU press release. Apple’s ATT (App Tracking Transparency) seems to have essentially done the same thing, without the legal delay, and around the world.
unique link to this extract


Financial models on climate risk ‘implausible’, say actuaries • Financial Times

Camilla Hodgson:

»

Financial institutions often did not understand the models they were using to predict the economic cost of climate change and were underestimating the risks of temperature rises, research led by a professional body of actuaries shows.

Many of the results emerging from the models were “implausible,” with a serious “disconnect” between climate scientists, economists, the people building the models and the financial institutions using them, a report by the Institute and Faculty of Actuaries and the University of Exeter finds.

Companies are increasingly required to report on the climate-related risks they face, using mathematical models to estimate how resilient their assets and businesses might be at different levels of warming.

The International Sustainability Standards Board last week launched long-awaited guidance for companies to inform investors about sustainability-related risks, including the climate scenarios chosen in their calculations.

Countries including the UK and Japan have said they plan to integrate these standards into their reporting rules.

Companies will also have to report the full scope of their emissions, including those from their supply chains, from the second year they begin to report under the guidelines due to come into effect in 2024.

…Some models were likely to have “limited use as they do not adequately communicate the level of risk we are likely to face if we fail to decarbonise quickly enough,” the paper released on Tuesday said.

«

unique link to this extract


Revealed: UK plans to drop flagship £11.6bn climate pledge • The Guardian

Helena Horton and Patrick Greenfield:

»

The government is drawing up plans to drop the UK’s flagship £11.6bn climate and nature funding pledge, the Guardian can reveal, with the prime minster accused of betraying populations most vulnerable to global heating.

The disclosure provoked fury from former ministers and representatives of vulnerable countries, who accused Rishi Sunak of making false promises.

A leaked briefing note to ministers, given to the Foreign Office and seen by the Guardian, lays out reasons for dropping the UK’s contribution to meeting the global $100bn (£78.6bn) a year commitment to developing countries.

It says: “Our commitment to double our international climate finance to £11.6bn was made in 2019, when we were still at 0.7 [% of GDP spent on international aid] and pre-Covid.” It adds that to meet it by the deadline would be a “huge challenge” because of new pressures, including help for Ukraine being included in the aid budget.

To meet the £11.6bn target by 2026, government officials have calculated that it would have to spend 83% of the Foreign Office’s official development assistance budget on the international climate fund. Civil servants said in the leaked document that this “would squeeze out room for other commitments such as humanitarian and women and girls”.

…The projects funded include building renewable energy, helping create low-pollution transport and protecting forests in sensitive areas around the globe.

Former Foreign Office minister Zac Goldsmith, who resigned last week citing what he termed Sunak’s “apathy” towards the environment, said this would “shred” the UK’s international reputation.

«

Alok Sharma, the Conservative MP who negotiated at the COP summit(s), also expressed disquiet. Astonishing lack of sway they have. Meanwhile the grousers who complain it’s all too expensive get airtime.
unique link to this extract


Are extraterrestrials extra patriotic? • The Economist

»

According to the National UFO Reporting Centre (NUFORC), an American non-profit organisation that has collected reports of unidentified flying objects since 1974, UFO sightings [in the US] tend to spike on July 4th. Between 1995 and 2018, around 2% of all sightings recorded by NUFORC fell on this date; seven times more than would be expected by chance. What explains this strange phenomenon?

Hollywood may be partly to blame. In the two years before the release of the Will Smith flick, NUFORC recorded an average of seven UFO sightings on July 4th (eight in 1995 and six in 1996). In 1997, a year after aliens burst onto the big screen, there were 74—more than ten times as many. Traditions of the July 4th holiday may also help explain the spike. Independence Day is typically spent outdoors. Heavy alcohol use is not uncommon. Intoxication may cause some to confuse celebratory fireworks with alien aircraft.

«

Makes sense to me.
unique link to this extract


No, the good economic data isn’t being faked • The New York Times

Paul Krugman:

»

During the Obama years there was a large faction of “inflation truthers,” who insisted that deficit spending and monetary expansion must surely be causing runaway inflation, and that if official numbers failed to match that prediction it was only because the government was cooking the books.

With inflation falling rapidly over the past year, we’ve seen some resurgence of inflation trutherism. But the more notable development has been the emergence of what we might call recession truthers — a significant faction that seems frustrated by the Biden economy’s refusal, at least so far, to enter the recession they have repeatedly predicted or insisted is already underway.

Now, there are some sociological differences between the old inflation truthers and the new recession truthers. The former group tended to be old-school reactionaries still pining for a return to the gold standard. The new group is dominated by tech bros, billionaires who imagine themselves focused on the future rather than the golden past, more likely to be crypto cultists than gold bugs. [Essentially the same thing – Overspill Ed]

…You might have expected technology billionaires to be well-informed about the world — someone like Musk could, if he chose, easily maintain a large research department for his personal edification. (The annual budget for the whole Bureau of Labor Statistics is less than $700m.) Yet they are often, in practice, easy marks for grifters and con men.

…So why do we see tech bros indulging in conspiracy theories, often citing random Twitter accounts to justify their views?

The answer, I believe, is that technology billionaires are especially susceptible to the belief that they’re uniquely brilliant, able to instantly master any subject, from Covid to the war in Ukraine. They could afford to hire experts to brief them on world affairs, but that would only work if they were willing to listen when the experts told them things they didn’t want to hear. So what happens instead, all too often, is that they go down the rabbit hole: Their belief in their own genius makes them highly gullible, easy marks for grifters claiming that the experts are all wrong.

«

David Sacks in particular completely demonstrates how correct Krugman is.
unique link to this extract


Fake journalist profiles used to launch Bournemouth Observer • Hold The Front Page

David Sharman:

»

Fake journalist profiles have been used to launch a new website purporting to cover local news in a UK town, an HTFP investigation has found.

Photos taken from a stock picture archive were used by the Bournemouth Observer, which claims to be a new independent title serving Bournemouth, Poole and Christchurch, to illustrate a series of profiles of its journalists.

The Observer, which also offers a range of advertising opportunities, had initially contacted HTFP about a potential directory listing on our site, but we decided to investigate after Paul Giles, a representative of the title, refused to provide basic details about the backgrounds of the journalists listed or even confirm they were real people.

Some of the Observer’s content has also raised questions about whether AI is being used to produce its copy after police failed to find any record of two incidents reported by the site.

A ‘Meet the Team’ page on the website, which has now been deleted following our enquiries, listed 11 members of staff with photos and biographies, but we cross-referenced the headshots with an online reverse image search tool and discovered that all 11 pictures were stock images.

At least seven of the images originated from the same stock photo archive – istockphoto.com

The biography for David Roberts, described as the Observer’s “esteemed editor” with “a career in journalism that spans decades”, also contained inconsistencies.

It claimed he has been “lending his expertise and leadership to the newspaper for several years”, despite the Observer only being launched last month as an online-only title.

«

The site was only registered on 30 May, and “registrant contact details [are] waiting to be checked” by Nominet. Feels like someone who is trying to use ChatGPT to write stories and just made up a ton of journalist name and profiles.
unique link to this extract


UK universities draw up guiding principles on generative AI • The Guardian

Sally Weale:

»

While once there was talk of banning software like ChatGPT within education to prevent cheating, the guidance says students should be taught to use AI appropriately in their studies, while also making them aware of the risks of plagiarism, bias and inaccuracy in generative AI.

Staff will also have to be trained so they are equipped to help students, many of whom are already using ChatGPT in their assignments. New ways of assessing students are likely to emerge to reduce the risk of cheating.

…The five guiding principles state that
• universities will support both students and staff to become AI literate;
• staff should be equipped to help students to use generative AI tools appropriately;
• the sector will adapt teaching and assessment to incorporate the “ethical” use of AI and ensure equal access to it
• universities will ensure academic integrity is upheld; and
• [universities will] share best practice as the technology evolves.

«

Good that they’re taking this seriously. And that’s not a bad set of principles.
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 comment

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