Stanford and the ethical dilemma of Silicon Valley’s next generation

This is a great insight into one of Silicon Valley’s most important brainbox factories, where current students, who once dreamed of landing at Google or Facebook, are grappling with the tech industry’s vilification. The Ringer’s Victor Luckerson:

As tech comes to dominate an ever-expanding portion of our daily lives, Stanford’s role as an educator of the industry’s engineers and a financier of its startups grows increasingly important. The school may not be responsible for creating our digital world, but it trains the architects. And right now, students are weighing tough decisions about how they plan to make a living in a world that was clearly constructed the wrong way. “To me it seemed super empowering that a line of code that I wrote could be used by millions of people the next day,” says Matthew Sun, a junior majoring in computer science and public policy, who helped organize the Theranos event. “Now we’re realizing that’s maybe not always a good thing.”

A report last year by the Wall Street Journal looked at the issue of employee retention and recruitment in the wake of a torrid year for Facebook’s reputation. It found that while existing employees were mostly happy (partly, the piece speculated, thanks to a stock price that has remained strong), but that there were some signs getting new people on board was becoming more difficult:

Since the disclosures, more candidates for jobs in some units at Facebook have withdrawn from consideration than during any other period in memory, according to a person familiar with the company’s recruiting. Mr. Zuckerberg said in a recent company meeting that the Cambridge Analytica flap didn’t seem to be deterring job applicants broadly across the company, according to a person familiar with the matter.

Instagram bans self-harm images

Less than a month ago, BBC journalist Angus Crawford reported on the absolutely heartbreaking story of Molly Russell, a 14-year-old girl who took her own life. Her father said he felt Instagram, which hosted images of self-harm, was partly responsible for what happened.

Today, Instagram boss Adam Mosseri announced the site will no longer allow (most) pictures depicting self-harm:

Adam Mosseri said Instagram was trying to balance “the need to act now and the need to act responsibly”.

He added the site was “not where we need to be on the issues of self-harm and suicide”.

When asked by the BBC’s Angus Crawford when the images would be removed, Mr Mosseri replied: “As quickly as we can, responsibly.”

Digital minister Margot James told BBC Radio 4’s PM programme the government would “have to keep the situation very closely under review to make sure that these commitments are made real – and as swiftly as possible”.

That last paragraph is especially important. Goes without saying that we’ll be keeping a very close eye on whether Instagram lives up to this promise.

In interviewing Mosseri, Crawford didn’t mess around:

When asked if he would resign if graphic self-harm content was still on the platform in six months, Mr Mosseri, 36, said: “I will certainly have a long thought about how well I am doing in the role that I’m in.”

On Spotify’s ability to revolutionise the profitability of podcasts

Verrrrryyy interesting deal by Spotify, confirmed on Wednesday, to acquire Gimlet and Anchor – two promising podcasting companies.

For BBC News I offered this analysis on how Spotify can inject some steroids into the podcast business, though not without risks:

What’s particularly powerful is Spotify’s ability to match users with podcasts the same way it does (spookily) well with new music. There are 16,000 different “signals”, the company says, that help it do that. Those same signals will mean Spotify bring the earning potential for podcasts to the next level: if you listen to a health and fitness playlist, you’re likely to respond well to podcast ads offering fitness gear, perhaps.

But on the other hand, I realise the future I’m describing here sounds rather awful. Imagine having a podcast on in the car, with your friends judging you based on the advertising that is being played out. “If you’re suffering from irritable bowel syndrome, you should try…”


Spotify typically takes its sweet time in rolling out new features – though that’s surely not a bad thing in this instance. Getting this right is going to be hard.

Using AI to locate victims of sex trafficking

From The Register:

A group of researchers from George Washington University, Temple University, and Adobe in the US have built a large dataset containing over a million images from 50,000 hotels across different countries. They hope their public Hotels-50K dataset will help developers train neural networks that can spot where a victim may be in seconds, judging from the background of their online ad.

A room’s decor may indicate its general vicinity, based on the hotel it is likely to be in. Curtains, wallpaper, bedspreads, and so on, can be analyzed to narrow down victims to particular chains and locations.

Very interesting work, though not something that should be undertaken without considering what some of the unintended consequences may be. Most obviously, identifying locations in this way may lead to pimps going to darker lengths to hide these women.

And this recent report should give added pause for thought:

Violent crime is way down in San Francisco, according to the latest police statistics. But one major category is bucking the trend: police recorded a 170 percent jump in reports of human trafficking in 2018.

The huge spike appears to be connected to the federal shutdown of sex-for-sale websites. The goal of shutting them down was to curb human trafficking. Instead, it seems to have had the opposite effect.

What can we learn from this? Measures to help women in the sex trade tend to be more successful when people listen to the women.

Rosalind Brewer, formerly of Starbucks, becomes Amazon’s first black board member

USA Today on the appointment of Rosalind Brewer at Amazon:

The appointment comes after Amazon caved to pressure from shareholders, employees and the Congressional Black Caucus in May and agreed to adopt a shareholder proposal to increase the diversity of its board. The proposal required Amazon to consider women and minorities for board openings, similar to the National Football League’s Rooney Rule, which has increased the diversity of coaching staff by mandating that teams interview at least one minority candidate.

She will become the 10th member and fourth woman on the board.

Attacking a pay wall that hides public court filings – NYT

Encouraging reading in the NYT this morning. A report details how there’s a fight brewing against Pacer, the atrocious and expensive US court records system:

The costs of storing and transmitting data have plunged, approaching zero. By one estimate, the actual cost of retrieving court documents, including secure storage, is about one half of one ten-thousandth of a penny per page. But the federal judiciary charges a dime a page to use its service, called Pacer (for Public Access to Court Electronic Records).

AI music is here, and it’s better than you think

I’m a big fan of stories about artificial intelligence that focuses on what technology can do, rather than obsessing over what it can’t.

A good example of this today is this piece from Stuart Dredge on AI-generated music, posted on Medium:

Can an A.I. create original music? Absolutely. Can it create original music better than a human can? Well, it depends which human you’re comparing the A.I.’s music to, for a start.

Human-created music already spans everything from the sublime to the unlistenable. While an A.I. may not be able to out-Adele Adele (or Aretha Franklin, or Joni Mitchell) with a timeless song and performance, it can compose a compelling melody for a YouTube video, mobile game, or elevator journey faster, cheaper, and almost as well as a human equivalent. In these scenarios, it’s often the “faster” and “cheaper” parts that matter most to whoever’s paying.

This demo, from Australian start-up Popgun, shows the current capabilities of AI to create competent, if not altogether inspiring, music:

As one of those people who does a lot of video editing to music, I do wonder if a future feature might be to get AI to generate a soundtrack based on my shots. A boy can dream.

Updated: Fact-checkers end work with Facebook

Another update: 5:03pm PT: The Associated Press has also decided not to renew its contract. Wuh-oh.

Statement: “AP is not currently doing fact-checking work for Facebook. AP constantly evaluates how to best deploy its fact-checking resources, and that includes ongoing conversations with Facebook about opportunities to do important fact-checking work on its platform.”

Update, 12:44pm PT: Read my report on this @ BBC News

Fact checking service Snopes has said today it was ending its partnership with Facebook. In a statement, this paragraph stands out. So much potentially between the lines here:

At this time we are evaluating the ramifications and costs of providing third-party fact-checking services, and we want to determine with certainty that our efforts to aid any particular platform are a net positive for our online community, publication, and staff.

Was it too expensive because of the workload? Was it having any effect?

Facebook’s statement:

We value the work that Snopes has done, and respect their decision as an independent business. Fighting misinformation takes a multi-pronged approach from across the industry. We are committed to fighting this through many tactics, and the work that third-party fact-checkers do is a valued and important piece of this effort. We have strong relationships with 34 fact-checking partners around the world who fact-check content in 16 languages, and we plan to expand the program this year by adding new partners and languages.

According to Snopes’ financial disclosures, Facebook paid them $100,000 for their work in 2017 (after a brief period doing it for nothing). Snopes hasn’t yet put out its financials for 2018. One can predict they got more money, but it will be telling to find out how much.

Even in 2017, $100,000 for this work seems low – that’s less than the average salary for one member of Facebook’s staff.

I’m reminded of this piece in the Guardian at the end of last year:

Journalists working as factcheckers for Facebook have pushed to end a controversial media partnership with the social network, saying the company has ignored their concerns and failed to use their expertise to combat misinformation.

Current and former Facebook factcheckers told the Guardian that the tech platform’s collaboration with outside reporters has produced minimal results and that they’ve lost trust in Facebook, which has repeatedly refused to release meaningful data about the impacts of their work. Some said Facebook’s hiring of a PR firm that used an antisemitic narrative to discredit critics – fueling the same kind of propaganda factcheckers regularly debunk – should be a deal-breaker.

‘Fake news about journalism’ can only be solved with transparency – but how?

Financial Times columnist Simon Kuper does his best to address some of the myths of modern journalism, albeit in a manner which makes it highly unlikely anyone who needs to read it actually will.

He outlines four main complaints/misconceptions. This is one of them:

“You are out of touch with ordinary people.” That’s becoming less true. Brexit and Trump’s election shocked media into trying to reconnect with ordinary folk — especially white folk. Hence the new American journalistic genre of “Trump safaris” (visits to white working-class towns), while just before the Brexit referendum the BBC decided to fund 150 new local reporters to cover local democracy around Britain.

I agree with him that the accusation is becoming less true. But, I disagree with his reasons why. The “Trump safaris” (in the UK you might call it a “Northern Expedition”) are routinely a stain on good reporting, bringing little more than a gawp at the poor and, through absolutely no fault of their own, less educated.

That aside, this trend – journalists showing how they work – is one that should gain momentum.

Anyone who has spent a meaningful amount of time in a good newsroom will know the lengths journalists go to do their work in an honest, open and passionate way. What we need to do is find out how we share that, continuously, with the public. Documentaries like Showtime’s Fourth Estate give a glimpse at political reporting in the US, but such insights are too infrequent.

It’s been acceptable, in the past, for journalists to merely show the end result of their hard work. What we need to work out is how to show our working, without compromising our integrity – and the safety of those brave enough to come to us.

What does Alexa bring to Amazon?

Amazon’s had the greatest success in the voice assistant race so far, surprising many – including Amazon, you’d imagine.

I have one in my kitchen, used for playing music, a couple of radio stations, the odd podcast (though I find the lack of controls frustrating) and asking the weather before I head out.

All very clever, but how is this of any use to Amazon? That’s the question that Benedict Evans tackles here:

[W]e’ve had two statistics recently from Amazon about Alexa: that 100m units have been sold, and that Amazon has 10,000 people working on it. This is vastly more people than one would normally expect to get the device made and the software working, especially given that the Echos themselves are produced by contract manufacturers. Rather, this is about experimentation and iteration: “what can we do with this and how can we expand the use case?” 

I don’t know the answer to this, but looking forward to seeing Amazon trying to figure it out. There has to be a strategy beyond sticking Alexa in absolutely everything. Or maybe there isn’t?