Start Up No.1362: how Google search got so closed, US lets off Twitter Saudi insiders, marathon masks, Trump tweets deleted, and more

Want a better picture from your webcam? Camo, a British app, lets you use your iPhone camera for Mac apps CC-licensed photo by Brett Renfer 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. Clearly. I’m @charlesarthur on Twitter. Observations and links welcome.

Google’s top search result? Surprise! It’s Google • The Markup

Adrianne Jeffries and Leon Yin:


A trending search in our data for “myocardial infarction” shows how Google has piled up its products at the top. It returned:

• Google’s dictionary definition
• A “people also ask” box that expanded to answer related questions without leaving the search results page
• A “knowledge panel,” which is an abridged encyclopedia entry with various links
• And a “related conditions” carousel leading to various new Google searches for other diseases

All of these appeared before search results by WebMD, Harvard University, and Medscape. In fact, a user would have to scroll nearly halfway down the page—about 42 percent—before reaching the first “organic” result in that search.

Google’s decision to place its products above competitors’ and to present “answers” on the search page has led to lawsuits and regulatory fines. A number of websites said it killed their revenues—and their companies. Founders of both innovative startups and companies that had been around for a decade or more told The Markup that once Google started placing its product first, they didn’t stand a chance.  

Travel research firm Skift wrote in November that the entire online travel industry is suffering. “The fact that Google is leveraging its dominance as a search engine into taking market share away from travel competitors is no longer even debatable.”


There’s a supporting article about how they did their research. It feels as though Google has decided not to leave it to chance any more: keep people on the site to show them ads.

Do read through to the point where it shows that Google doesn’t always (often?) offer the best prices for flights, either. That’s the danger of shopping monopolies: you might not see what is being hidden from you.
unique link to this extract

Ex-Twitter workers win US case dismissal over Saudi hacks • Bloomberg via MSN

Clare Roth and Peter Blumberg:


The US sought to dismiss charges it brought late last year against two former Twitter Inc. employees and a Saudi national for allegedly helping Riyadh spy on dissidents who use the social network.

Prosecutors in San Francisco on Tuesday asked for a judge’s permission to drop the charges. The two-page filing doesn’t offer a reason but specifies that the dismissal would be “without prejudice,” meaning the government could file new charges.

The two former Twitter employees, Ahmad Abouammo and Ali Alzabarah, were accused of feeding the Saudi government information about Twitter users critical of it. They were recruited by a Saudi named Ahmed Almutairi, who lives in the kingdom and has worked for the royal family’s social media company, according to prosecutors.

All three were charged with acting as illegal foreign agents. Of the three, only Abouammo, a U.S. citizen, is in custody. He has pleaded not guilty.

Twitter, the Saudi Embassy, a lawyer for Abouammo and the U.S. attorney in San Francisco didn’t immediately respond to calls and emails seeking comment on the prosecutor’s request for dismissal.


What possible reason would there be to withdraw these charges? The suspicion is that this is some corrupt deal sewn up by the US DOJ with Saudi Arabia. Four years ago, that wouldn’t have been countenanced. Now, it’s the first suggestion.
unique link to this extract

The rise of synthetic audio deepfakes • Nisos Security

Robert Volkert, VP Threat Investigations and Dev Badlu, VP Technology at Nisos:


Audio deepfakes are the new frontier for business compromise schemes and are becoming more common pathways for criminals to deceptively gain access to corporate funds. Nisos recently investigated and obtained an original attempted deepfake synthetic audio used in a fraud attempt against a technology company. The deepfake took the form of a voicemail message from the company’s purported CEO, asking an employee to call back to “finalize an urgent business deal.” The recipient immediately thought it suspicious and did not contact the number, instead referring it to their legal department, and as a result the attack was not successful.

Nisos investigated the phone number the would-be attacker used and determined it was a VOIP service with no owner registration information. It was likely simply acquired and used as a “burner” for this fraud attempt only. While there was no actual voicemail message associated with the number, we made no attempt for live contact with the owner of the phone number for legal reasons.

…The most famous use of deep fake synthetic audio technology in criminal fraud was a September 2019 incident involving a British energy company. The criminals reportedly used voice-mimicking software to imitate the British executive’s speech and trick his subordinate into sending hundreds of thousands of dollars to a secret account.

The managing director of this company, believing his boss was on the phone, followed orders to wire more than $240,000 to an account in Hungary.1


So the threat from audio deepfakes is really to business rather than to politics. So far, anyhow.
unique link to this extract

Making news and enlightening audiences: BBC’s flagship news show in the pandemic • Reuters Institute for the Study of Journalism

Sarah Sands is leaving as editor of the BBC’s Today programme after three years, and last December was facing a boycott by government ministers who thought they were going to teach the BBC a lesson; they wouldn’t appear on the show:


if there is a problem with the show, it’s down to me.

And having no ministers is a problem.  

I said I would not beg [for forgiveness, because the programme wasn’t at fault], and I didn’t. But I did go to Downing Street to see if we could find a way through. Serious issues were on the agenda – floods, a big decision on HS2, Huawei. The Today programme seemed the right place to talk about them. Downing Street wondered what was in it for the government.

On January 29, Chinese nationals at a hotel in York [were] reported to have fallen ill, the first coronavirus cases on British soil. 

Over the following month, coverage of the crisis increased by the day. We talked to doctors, to epidemiologists, WHO officials, to the brightest minds we could find. They came on willingly and shared everything they knew. Everyone seemed happy to answer intelligent questions for an intelligent audience. Everyone, except the government.    

And something really interested happened during those weeks.


What if the pandemic hadn’t intervened? I think the result would have been the same. The government would discover that it needed to be heard more than it could use social media.
unique link to this extract

Facebook’s ‘red team’ hacks its own AI programs • WIRED

Tom Simonite:


Deepfake technology is becoming easier to access and has been used for targeted harassment. When Canton’s group formed last year, researchers had begun to publish ideas for how to automatically filter out deepfakes. But he found some results suspicious. “There was no way to measure progress,” he says. “Some people were reporting 99% accuracy, and we were like ‘That is not true.’”

Facebook’s AI red team launched a project called the Deepfakes Detection Challenge to spur advances in detecting AI-generated videos. It paid 4,000 actors to star in videos featuring a variety of genders, skin tones, and ages. After Facebook engineers turned some of the clips into deepfakes by swapping people’s faces around, developers were challenged to create software that could spot the simulacra.

The results, released last month, show that the best algorithm could spot deepfakes not in Facebook’s collection only 65% of the time. That suggests Facebook isn’t likely to be able to reliably detect deepfakes soon. “It’s a really hard problem, and it’s not solved,” Canton says.

Canton’s team is now examining the robustness of Facebook’s misinformation detectors and political ad classifiers. “We’re trying to think very broadly about the pressing problems in the upcoming elections,” he says.

Most companies using AI in their business don’t have to worry as Facebook does about being accused of skewing a presidential election. But Ram Shankar Siva Kumar, who works on AI security at Microsoft, says they should still worry about people messing with their AI models. He contributed to a paper published in March that found 22 of 25 companies queried did not secure their AI systems at all. “The bulk of security analysts are still wrapping their head around machine learning,” he says. “Phishing and malware on the box is still their main thing.”


unique link to this extract

Camo – Use your phone as a pro webcam, free • Reincubate



Look amazing on video calls. Use your iPhone or iPad as a pro webcam and get powerful effects and adjustments for Zoom, Meet, and more.


A British software company with the answer to Joanna Stern’s (and everyone else’s) prayers: use your iPhone as the webcam while you use your Mac. (As recommended on Benedict Evans’s newsletter.)

There are a growing number of iPhone/iPad + Mac app pairings – Duet is another, to make an iPad work as a second screen for a Mac. Others?
unique link to this extract

Their businesses went virtual. Then Apple wanted a cut • The New York Times

Jack Nicas and David McCabe:


ClassPass built its business on helping people book exercise classes at local gyms. So when the pandemic forced gyms across the United States to close, the company shifted to virtual classes.

Then ClassPass received a concerning message from Apple. Because the classes it sold on its iPhone app were now virtual, Apple said it was entitled to 30% of the sales, up from no fee previously, according to a person close to ClassPass who spoke on the condition of anonymity for fear of upsetting Apple. The iPhone maker said it was merely enforcing a decade-old rule.

Airbnb experienced similar demands from Apple after it began an “online experiences” business that offered virtual cooking classes, meditation sessions and drag-queen shows, augmenting the in-person experiences it started selling in 2016, according to two people familiar with the issues.

Airbnb discussed Apple’s demands with House lawmakers’ offices that are investigating how Apple controls its App Store, according to three people who spoke on the condition of anonymity to discuss private conversations. Those lawmakers are now considering Apple’s efforts to collect a commission from Airbnb and ClassPass as part of their yearlong antitrust inquiry into the biggest tech companies, according to a person with knowledge of their investigation.

…With gyms shut down, ClassPass dropped its typical commission on virtual classes, passing along 100% of sales to gyms, the person close to the company said. That meant Apple would have taken its cut from hundreds of struggling independent fitness centers, yoga studios and boxing gyms.

Apple said that with Airbnb and ClassPass, it was not trying to generate revenue — though that is a side effect — but instead was trying to enforce a rule that has been in place since it first published its app guidelines in 2010.


That’s going to make for an interesting session when Tim Cook gets grilled by the US House Antitrust Committee.
unique link to this extract

Sen. Josh Hawley wants to strip legal protections from sites with targeted ads • The Verge

Adi Robertson:


Sen. Josh Hawley (Republican, Missouri) has introduced the latest of several bills designed to weaken a key online legal shield. The Behavioral Advertising Decisions Are Downgrading Services (or BAD ADS) Act would remove protections under Section 230 of the Communications Decency Act for large web services that display ads based on “personal traits of the user” or a user’s previous online behavior. This is defined as “behavioral advertising” and does not include targeting based on users’ locations or the content of the site they’re on.

Section 230 shields websites from legal liability for user-created content. Unlike several previous bills, including ones sponsored by Hawley, the BAD ADS Act doesn’t appear to address any specific critiques of Section 230. It’s seemingly an anti-targeted ads bill that threatens companies with the loss of an unrelated legal protection instead of monetary fines. Hawley has previously introduced a bill that would create a Do Not Call list equivalent for targeted advertising, and he’s proposed banning “addictive” design features like endless scrolling on social networking apps.

In a statement, Hawley said that “manipulative ads are not what Congress had in mind when passing Section 230,” although he did not elaborate on a relationship between the two topics.


The trouble with Hawley’s plan is that stripping s230 would leave such companies liable for absolutely everything, so they’d just block everything except the most vanilla content, and certainly anything that might get them co-sued. It’s a self-defeating move: why wouldn’t someone sue Hawley for things he tweets or posts on Facebook causing some sort of harm?

Here’s my suggestion (pass it on): put a ceiling on the size of social networks. 100 million, 200 million? The specific danger comes from size, not from particularly what they do. Let the right-wing nutjobs be on Gab or Parler. Let the others be on Mastodon,, whatever. But put the limit on how many they can reach.
unique link to this extract

Two Donald Trump tweets deleted by Twitter overnight •

Steph Bazzle:


Twitter has started acting when a tweet from Donald Trump violates their site rules. Over Monday night, two tweets from the president were removed. Unlike some of the recent responses from the site, this time they didn’t include a statement about the content. However, the tweets in question are archived, and the content may be a clue.

According to Factbase, one of the tweets was about COVID-19. It was a retweet, and while the link to the article is no longer visible, the text describes the claims of one Dr. Vladimir Zelenko. “I have treated over 350 patients [using hydroxychloroquine] with 100% success.”

Here’s the problem with this claim, as documented by the New York Times back in April (when Zelenko was already claiming 350 cases cured). Zelenko’s claims aren’t backed by evidence, the officials in his New York village, Kiryas Joel, have asked him to stop, saying that he’s exaggerating the outbreak in their community and inflating numbers by falsifying the number who became ill, and the numbers don’t reflect what scientific studies continue to find. Further, without sufficient testing, any suggestion for treating pre-symptomatic patients becomes moot.

Trump’s second deleted tweet was also a retweet. This one linked to an article by The Post Millenial. It claims that Garrett Foster, a protester killed in Austin, shot at a car five times before a driver fired back, killing him. That article has since been updated to include the following correction [saying that Foster, who was killed, was not the first shooter].


Interesting that we haven’t heard any screaming from Trump about this. (Related: his dim son – OK, Junior, I have to narrow it down – had his Twitter access limited for tweeting nonsense about hydroxychloroquine. So of course the right-wingers claimed it was election interference. Night, day, follow.)
unique link to this extract

A doctor ran 22 miles with a face mask on to debunk a ridiculous myth • BGR

Chris Smith:


Face mask protesters have made up several silly reasons to oppose the use of the simplest tool possible to reduce the risk of COVID-19 transmission. Some oppose the idea of being told what to do, suggesting face masks are about complying with the government. They are not. It’s about curbing the spread of the virus. Others claim the use of face masks is actually harmful, claiming they can reduce the flow of oxygen. That’s also false. Again, the use of face covers can reduce the spread of a pathogen, protecting both the person wearing the mask and others.

A few weeks ago, a doctor put on six surgical masks at the same time to prove it wouldn’t affect his breathing. He wore a pulse oximeter, a device that measures oxygen saturation, as proof. Unsurprisingly, the medical gadget confirmed his blood oxygen levels stayed within normal parameters. Another doctor performed an even more audacious task to dispel the hypoxia myth; he ran 22 miles and monitored his oxygen with the same type of device. The conclusion was identical: face masks do not reduce the flow of oxygen, even if you’re running and need a much higher intake of air to supply the increased oxygen needs of the muscles.

Dr. Tom Lawton from the Bradford Royal Infirmary in Yorkshire, England decided to run with a face mask on to fight misinformation and the spread of fake news about face masks. His oxygen levels never fell below 98% during the course of his run — any value of over 94% is considered normal.


unique link to this extract

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.