Joho the BlogJoho the Blog - Page 3 of 964 - Let's just see what happens

September 26, 2017

[liveblog][PAIR] Maya Gupta on controlling machine learning

At the PAIR symposium. Maya Gupta runs Glass Box at Google, which looks at black box issues. She is talking about how we can control machine learning to do what we want

NOTE: Live-blogging. Getting things wrong. Missing points. Omitting key information. Introducing artificial choppiness. Over-emphasizing small matters. Paraphrasing badly. Not running a spellpchecker. Mangling other people’s ideas and words. You are warned, people.

The core idea of machine learning are its role models, i.e., its training data. That’s the best way to control machine learning. She’s going to address by looking at the goals of controlling machine learning.

A simple example of monotinicity. Let’s say we’re tring to recommend nearby coffee shops. So we use data about the happiness of customers and distance from the shop. We can fit the model ot a linear model. Or we can fit it to a curve, which works better for nearby shops but goes wrong for distant shops. That’s fine for Tokyo but terrible for Montana because it’ll be sending people many miles away. A montonic example says we don’t want to do that. This controls ML to make it more useful. Conclusion: the best ML has the right examples and the right kinds of flexibility. [Hard to blog this without her graphics. Sorry.] See “Deep Lattice Networks for Learning Partial Monotonic Models,” NIPS 2017; it will soon by on the TensorFlow site.

“The best way to do things for practitioners is to work next to them”The best way to do things for practitioners is to work next to them.

A fairness goal: e.g., we want to make sure that accuracy in India is the same as accuracy in the US. So, add a constraint that says what accuracy levels we want. Math lets us do that.

Another fairness goal: the rate of positive classifications should be the same in India as in the US, e.g., rate of students being accepted to a college. In one example, there is an accuracy trade-off in order to get fairness. Her attitude: Just tell us what you want and we’ll do it

Fairness isn’t always relative. E.g., E.g., minimize classification errors differently for different regions. You can’t always get what you want, but you sometimes can or can get close. [paraphrase!] See

It can be hard to state what we want, but we can look at examples. E.g., someone hand-labels 100 examples. That’s not enough as training date, but we can train the system so that it classifies those 100 at something like 95% accuracy.

Sometimes you want to improve an existing ML system. You don’t want to retrain because you like the old results. So, you can add in a constraint such as keep the differences from the original classifications to less than 2%.

You can put all of the above together. See “Satisfying Real-World Goals with Dataset Constraints,” NIPS, 2016. Look for tools coming to TensorFlow.

Some caveats about this approach.

First, to get results that are the same for men and women, the data needs to come with labels. But sometimes there are privacy issues about that. “Can we make these fairness goals work without labels? ”Can we make these fairness goals work without labels? Research so far says the answer is messy. E.g., if we make ML more fair for gender (because you have gender labels), it may also make it fairer for race.

Second, this approach relies on categories, but individuals don’t always fit into categories. But, usually if you get things right on categories, it usually works out well in the blended examples.

Maya is an optimist about ML. “But we need more work on the steering wheel.” We’re not always sure we want to go with this technology. And we need more human-usable controls.

Comments Off on [liveblog][PAIR] Maya Gupta on controlling machine learning

[liveblog][PAIR] Hae Won Park on living with AI

At the PAIR conference, Hae Won Park of the MIT Media Lab is talking abiut personal social robots for home.

NOTE: Live-blogging. Getting things wrong. Missing points. Omitting key information. Introducing artificial choppiness. Over-emphasizing small matters. Paraphrasing badly. Not running a spellpchecker. Mangling other people’s ideas and words. You are warned, people.

Home robots no longer look like R2D2. She shows a 2014 Jibo video.

What sold people is the social dynamic: Jibo is engaged with family members.

She wants to talk about the effect of AI on people’s lives at home.

For example, Google Home changed her morning routine, how she purchases goods, and how she controls her home environment. She shows a 2008 robot called Autom, a weight management coach.

A studied showed that the robot kept people at it longer than using paper or a computer program, and people had the strongest “working alliance” with the robot. They also had emotional engagement with it, personalizing it, giving them names, etc. These users understand it’s just a machine. Why?

She shows a video of Leonardo, a social robot that exhibits bodily cues of emotion. We seem to share a mental model.

They studied how children tell stories and listen to each other. Jin Joo Lee developed a model in which the robot appears to be very attentive as the child tells a story. It notes cues about whether the speaker is engaged. The children were indeed engaged by this reactive behavior.

Researchers have found that social robots activate social thinking, lighting up the social thinking part of the brain. Social modeling occurs between humans and robots too.

Working with children aged 4-6, they studied “growth mindset”: the belief that you can get better if you try hard. Parents and teachers have been shown to affect this. They created a growth mindset robot that plays a game with the child. The robot encourages the child at times determined by a “Boltzmann Machine””>Boltzmann Machine. [Over my head.]

Their researc showed that playing puzzles with a growth-mindset robot fosters that mindset in children. For example, the children tried harder over time.

They also studied early literacy education using personalized robot tutors. In multiple studies of about 120 children. The robot, among other things, encourages the child to tell stories. Over four weeks, they found children more effectively learn vocabulary, and when the robot provided more expressive story telling (rather than speaking in an affect-less TTY voice) the children retained more and would mimic that expressiveness.

Now they’re studying fully automonmous storytelling robots. The robot uses the child’s responses to further engage the child. The children respond more, tell longer stories, and stayed engaged over longer periods across sessions.

We are headed toward a time when robots are more human-centered rather than task focused. So we need to think about making AI not just human-like but humanistic. We hope to make AI that make us better people.

Comments Off on [liveblog][PAIR] Hae Won Park on living with AI

[liveblog][PAIR] Karrie Karahalios

At the Google PAIR conference, Karrie Karahalios is going to talk about how people make sense of their world and lives online. (This is an information-rich talk, and Karrie talks quickly, so this post is extra special unreliable. Sorry. But she’s great. Google her work.)

NOTE: Live-blogging. Getting things wrong. Missing points. Omitting key information. Introducing artificial choppiness. Over-emphasizing small matters. Paraphrasing badly. Not running a spellpchecker. Mangling other people’s ideas and words. You are warned, people.

Today, she says, people want to understand how the information they see comes to them. Why does it vary? “Why do you get different answers depending on your wifi network? ”Why do you get different answers depending on your wifi network? These algorithms also affect our personal feeds, e.g., Instagram and Twitter; Twitter articulates it, but doesn’t tell you how it decides what you will see

In 2012, Christian Sandvig and [missed first name] Holbrook were wondering why they were getting odd personalized ads in their feeds. Most people were unaware that their feeds are curated: only 38% were aware of this in 2012. Thsoe who were aware became aware through “folk theories”: non-authoritative explanations that let them make sense of their feed. Four theories:

1. Personal engagement theory: If you like and click on someone, the more of that person you’ll see in your feed. Some people were liking their friends’ baby photos, but got tired of it.

2. Global population theory: If lots of people like, it will show up on more people’s feeds.

3. Narcissist: You’ll see more from people who are like you.

4. Format theory: Some types of things get shared more, e.g., photos or movies. But people didn’t get

Kempton studied thermostats in the 1980s. People either thought of it as a switch or feedback, or as a valve. He looked at their usage patterns. Regardless of which theory, they made it work for them.

She shows an Orbitz page that spits out flights. You see nothing under the hood. But someone found out that if you use a Mac, your prices were higher. People started using designs that shows the seams. So, Karrie’s group created a view that showed the feed and all the content from their network, which was three times bigger than what they saw. For many, this was like awakening from the Matrix. More important, they realized that their friends weren’t “liking” or commenting because the algorithm had kept their friends from seeing what they posted.

Another tool shows who you are seeing posts from and who you are not. This was upsetting for many people.

After going through this process people came up with new folk theories. E.g., they thought it must be FB’s wisdom in stripping out material that’s uninteresting one way or another. [paraphrasing].

They let them configure who they saw, which led many people to say that FB’s algorithm is actually pretty good; there was little to change.

Are these folk theories useful? Only two: personal engagement and control panel, because these let you do something. But there are poor tweaking tools.

How to embrace folk theories: 1. Algorithm probes, to poke and prod. “It would be great, Karrie says, to have open APIs so people could create tools”(It would be great to have open APIs so people could create tools. FB deprecated it.) 2. Seamful interfaces to geneate actionable folk theories. Tuning to revert of borrow?

Another control panel UI, built by Eric Gilbert, uses design to expose the algorithms.

She ends with a wuote form Richard Dyer: “All technolgoies are at once technical and also always social…”

Comments Off on [liveblog][PAIR] Karrie Karahalios

[liveblog][PAIR] Jess Holbrook

I’m at the PAIR conference at Google. Jess Holbrook is UX lead for AI. He’s talking about human-centered machine learning.

NOTE: Live-blogging. Getting things wrong. Missing points. Omitting key information. Introducing artificial choppiness. Over-emphasizing small matters. Paraphrasing badly. Not running a spellpchecker. Mangling other people’s ideas and words. You are warned, people.

“We want to put AI into the maker toolkit, to help you solve real problems.” One of the goals of this: “How do we democratize AI and change what it means to be an expert in this space?” He refers to a blog post he did with Josh Lovejoy about human-centered ML. He emphasizes that we are right at the beginning of figuring this stuff out.

Today, someone finds a data set, and finds a problem that that set could solve. You train a model and look at its performance, and decided if it’s good enough. And then you launch “The world’s first smart X. Next step: profit.” But what if you could do this in a human-centered way?

Human-centered design means: 1. Staying proximate. Know your users. 2. Inclusive divergence: reach out and bring in the right people. 3. Shared definition of success: what does it mean to be done? 4. Make early and often: lots of prototyping. 5. Iterate, test, throw it away.

So, what would a human-centered approach to ML look like? He gives some examples.

Instead of trying to find an application for data, human-centered ML finds a problem and then finds a data set appropriate for that problem. E.g., diagnosis plant diseases. Assemble tagged photos of plants. Or, use ML to personalize a “balancing spoon” for people with Parkinsons.

Today, we find bias in data sets after a problem is discoered. E.g., ProPublica’s article exposing the bias in ML recidivism predictions. Instead, proactively inspect for bias, as per JG’s prior talk.

Today, models personalize experiences, e.g., keyboards that adapt to you. With human-centered ML, people can personalize their models. E.g., someone here created a raccoon detector that uses images he himself took and uploaded, personalized to his particular pet raccoon.

Today, we have to centralize data to get results. “With human-centered ML we’d also have decentralized, federated learning”With human-centered ML we’d also have decentralized, federated learning, getting the benefits while maintaining privacy.

Today there’s a small group of ML experts. [The photo he shows are all white men, pointedly.] With human-centered ML, you get experts who have non-ML domain expertise, which leads to more makers. You can create more diverse, inclusive data sets.

Today, we have narrow training and testing. With human-centered ML, we’ll judge instead by how systems change people’s lives. E.g., ML for the blind to help them recognize things in their environment. Or real-time translation of signs.

Today, we do ML once. E.g., PicDescBot tweets out amusing misfires of image recognition. With human-centered ML we’ll combine ML and teaching. E.g., a human draws an example, and the neural net generates alternatives. In another example, ML improved on landscapes taken by StreetView, where it learned what is an improvement from a data set of professional photos. Google auto-suggest ML also learns from human input. He also shows a video of Simone Giertz, “Queen of the Shitty Robots.”

He references Amanda Case: “Expanding people’s definion of normal” is almost always a gradual process.

[The photo of his team is awesomely diverse.]

Comments Off on [liveblog][PAIR] Jess Holbrook

[liveblog] Google AI Conference

I am, surprisingly, at the first PAIR (People + AI Research) conference at Google, in Cambridge. There are about 100 people here, maybe half from Google. The official topic is: “How do humans and AI work together? How can AI benefit everyone?” I’ve already had three eye-opening conversations and the conference hasn’t even begun yet. (The conference seems admirably gender-balanced in audience and speakers.)

NOTE: Live-blogging. Getting things wrong. Missing points. Omitting key information. Introducing artificial choppiness. Over-emphasizing small matters. Paraphrasing badly. Not running a spellpchecker. Mangling other people’s ideas and words. You are warned, people.

The great Martin Wattenberg (half of Wattenberg – Fernanda Viéga) kicks it off, introducing John Giannandrea, a VP at Google in charge of AI, search, and more. “We’ve been putting a lot of effort into using inclusive data sets.”

John says that every vertical will affected by this. “It’s important to get the humanistic side of this right.” He says there are 1,300 languages spoken world wide, so if you want to reach everyone with tech, machine learning can help. Likewise with health care, e.g. diagnosing retinal problems caused by diabetes. Likewise with social media.

PAIR intends to use engineering and analysis to augment expert intelligence, i.e., professionals in their jobs, creative people, etc. And “how do we remain inclusive? How do we make sure this tech is available to everyone and isn’t used just by an elite?”

He’s going to talk about interpretability, controllability, and accessibility.

Interpretability. Google has replaced all of its language translation software with neural network-based AI. He shows an example of Hemingway translated into Japanese and then back into English. It’s excellent but still partially wrong. A visualization tool shows a cluster of three strings in three languages, showing that the system has clustered them together because they are translations of the same sentence. [I hope I’m getting this right.] Another example: a photo of integrated gradients hows that the system has identified a photo as a fire boat because of the streams of water coming from it. “We’re just getting started on this.” “We need to invest in tools to understand the models.”

Controllability. These systems learn from labeled data provided by humans. “We’ve been putting a lot of effort into using inclusive data sets.” He shows a tool that lets you visuallly inspect the data to see the facets present in them. He shows another example of identifying differences to build more robust models. “We had people worldwide draw sketches. E.g., draw a sketch of a chair.” In different cultures people draw different stick-figures of a chair. [See Eleanor Rosch on prototypes.] And you can build constraints into models, e.g., male and female. [I didn’t get this.]

Accessibility. Internal research from Youtube built a model for recommending videos. Initially it just looked at how many users watched it. You get better results if you look not just at the clicks but the lifetime usage by users. [Again, I didn’t get that accurately.]

Google open-sourced Tensor Flow, Google’s AI tool. “People have been using it from everything to to sort cucumbers, or to track the husbandry of cows.”People have been using it from everything to to sort cucumbers, or to track the husbandry of cows. Google would never have thought of this applications.

AutoML: learning to learn. Can we figure out how to enable ML to learn automatically. In one case, it looks at models to see if it can create more efficient ones. Google’s AIY lets DIY-ers build AI in a cardboard box, using Raspberry Pi. John also points to an Android app that composes music. Also, Google has worked with Geena Davis to create sw that can identify male and female characters in movies and track how long each speaks. It discovered that movies that have a strong female lead or co-lead do better financially.

He ends by emphasizing Google’s commitment to open sourcing its tools and research.



Fernanda and Martin talk about the importance of visualization. (If you are not familiar with their work, you are leading deprived lives.) When F&M got interested in ML, they talked with engineers. ““ML is very different. Maybe not as different as software is from hardware. But maybe. ”ML is very different. Maybe not as different as software is from hardware. But maybe. We’re just finding out.”

M&F also talked with artists at Google. He shows photos of imaginary people by Mike Tyka created by ML.

This tells us that AI is also about optimizing subjective factors. ML for everyone: Engineers, experts, lay users.

Fernanda says ML spreads across all of Google, and even across Alphabet. What does PAIR do? It publishes. It’s interdisciplinary. It does education. E.g., TensorFlow Playground: a visualization of a simple neural net used as an intro to ML. They opened sourced it, and the Net has taken it up. Also, a journal called aimed at explaining ML and visualization.

She “shamelessly” plugs deeplearn.js, tools for bringing AI to the browser. “Can we turn ML development into a fluid experience, available to everyone?”
What experiences might this unleash, she asks.

They are giving out faculty grants. And expanding the Brain residency for people interested in HCI and design…even in Cambridge (!).

Comments Off on [liveblog] Google AI Conference

September 19, 2017

[bkc] Hate speech on Facebook

I’m at a Very Special Harvard Berkman Klein Center for Internet & Society Tuesday luncheon featuring Monika Bickert, Facebook’s Head of Global Policy Management in conversation with Jonathan Zittrain. Monika is in charge of what types of content can be shared on FB, how advertisers and developer interact with the site, and FB’s response to terrorist content. [NOTE: I am typing quickly, getting things wrong, missing nuance, filtering through my own interests and biases, omitting what I can’t hear or parse, and not using a spelpchecker. TL;DR: Please do not assume that this is a reliable account.]

Monika: We have more than 2B users…

JZ: Including bots?

MB: Nope, verified. Billions of messages are posted every day.

[JZ posts some bullet points about MB’s career, which is awesome.]

JZ: Audience, would you want to see photos of abused dogs taken down. Assume they’re put up without context. [It sounds to me like more do not want it taken down.]

MB: The Guardian covered this. [Maybe here?] The useful part was it highlighted how much goes into the process of deciding these things. E.g., what counts as mutilation of an animal? The Guardian published what it said were FB’s standards, not all of which were.

MB: For user generated content there’s a set of standards that’s made public. When a comment is reported to FB, it goes to a FB content reviewer.

JZ: What does it take to be one of those? What does it pay?

MB: It’s not an existing field. Some have content-area expertise, e.g., terrorism. It’s not a minimum wage sort of job. It’s a difficult, serious job. People go through extensive training, and continuing training. Each reviewer is audited. They take quizzes from time to time. Our policies change constantly. We have something like a mini legislative session every two weeks to discuss proposed policy changes, considering internal suggestions, including international input, and external expert input as well, e.g., ACLU.

MB: About animal abuse: we consider context. Is it a protest against animal cruelty? After a natural disaster, you’ll see awful images. It gets very complicated. E.g., someone posts a photo of a bleeding body in Syria with no caption, or just “Wow.” What do we do?

JZ: This is worlds away from what lawyers learn about the First Amendment.

MB: Yes, we’re a private company so the Amendment doesn’t apply. Behind our rules is the idea that “You don’t have to agree with the content, but you should feel safe”FB should be a place where people feel safe connecting and expressing themselves. You don’t have to agree with the content, but you should feel safe.

JZ: Hate speech was defined as an attack against a protected category…

MB: We don’t allow hate speech, but no two people define it the same way. For us, it’s hate speech if you are attacking a person or a group of people based upon a protected characteristic — race, gender, gender identification, etc. —. Sounds easy in concept, but applying it is hard. Our rule is if I say something about a protected category and it’s an attack, we’d consider it hate speech and remove it.

JZ: The Guardian said that in training there’s a quiz. Q: Who do we protect: Women drivers, black children, or white men? A: White men.

MB: Not our policy any more. Our policy was that if there’s another characteristic beside the protected category, it’s not hate speech. So, attacking black children was ok but not white men, because of the inclusion of “children.” But we’ve changed that. Now we would consider attacks on women drivers and black children as hate speech. But when you introduce other characteristics such as profession, it’s harder. We’re evaluating and testing policies now. We try marking content and doing a blind test to see how it affects outcomes. [I don’t understand that. Sorry.]

JZ: Should the internal policy be made public?

MB: I’d be in favor of it. Making the training decks transparent would also be useful. It’s easier if you make clear where the line is.

JZ: Do protected categories shift?

MB: Yes, generally. I’ve been at FB for 5.5 yrs, in this are for 4 yrs. Overall, we’ve gotten more restrictive. Sometimes something becomes a topic of news and we want to make sure people can discuss it.

JZ: Didi Delgado’s post “all white people are racist” was deleted. But it would have been deleted if had said that all black people are racist, right?

MB: Yes. “If it’s a protected characteristic, we’ll protect it”If it’s a protected characteristic, we’ll protect it. [Ah, if only life were that symmetrical.]

JZL How about calls to violence, e.g., “Someone shoot Trump/Hillary”? If you think it should be taken down. [Sounds like most would let it stand.]

JZ: How about “Kick a person with red hair.” [most let it stand]

JZ: “How about: To snap a bitch’s neck, make sure to apply all your pressure to the middle of her throat.” [most let it stand][fuck, that’s hard to see up on the screen.]

JZ: “Let’s beat up the fat kids.” [most let it stand]

JZ: “#stab and become the fear of the Zionist” [most take it down]

MB: We don’t allow credible calls for violence.

JZ: Suppose I, a non-public figure, posted “Post one more insult and I’ll kill you.”

MB: We’d take that down. We also look at the degree of violence. Beating up and kicking might not rise to the standard. Snapping someone’s neck would be taken down, although if it were purely instructions on how to do something, we’d leave it up. “Zionist” is often associated with hate speech, and stabbing is serious, so we’d take them down. We leave room for aspirational statements wishing some bad thing would happen. “Someone should shoot them all” we’d count as a call to violence. We also look for specifity, as in “Let’s kill JZ. He leaves work at 3.” We also look at the vulnerability of people; if it’s a dangerous situation,
we’ll tend to treat all such things as calls to violence, [These are tough questions, but I’m not aligned with FB’s decisions on this.]

JZ: How long does someone spend reviewing this stuff?

MB: Some is easy. Nudity is nudity, although we let breast cancer photos through. But a beheading video is prohibited no matter what the context. Profiles can be very hard to evaluate. E.g., is this person a terrorist?

JZ: Given the importance of FB, does it seem right that these decisions reside with FB as a commercial entity. Or is there some other source that would actually be a relief?

MB: “We’re not making these decisions in a silo”We’re not making these decisions in a silo. We reach out for opinions outside of the company. We have Safety Advisory Board, a Global Safety Network [got that wrong, I think], etc.

JZ: These decisions are global? If I insult the Thai King…

MB: That doesn’t violate our global community standard. We have a group of academics around the world, and people on our team, who are counter-terrorism experts. It’s very much a conversation with the community.

JZ: FB requires real names, which can be a form of self-doxxing. Is the Real Name policy going to evolve?

MB: It’s evolved a little about what counts as their real name, i.e., the name people call you as opposed to what’s on your drivers license. Using your real name has always been a cornerstone of FB. A quinessential element of FB.

JZ: You don’t force disambiguation among all the Robert Smiths…

MB: When you communicate with people you know, you know you know them. “We don’t want people to be communicating with people who are not who you think they are”We don’t want people to be communicating with people who are not who you think they are. When you share something on FB, it’s not public or private. You can choose which groups you want to share it with, so you know who will see it. That’s part of the real name policy as well.

MB: We have our community standards. Sometimes we get requests from countries to remove violations of their law, e.g., insults to the King of Thailand. If we get such a request, if it doesn’t violate the standards, we look if the request is actually about real law in that country. Then we ask if it is political speech; if it is, to the extent possible, we’ll push back on those requests. E.g., Germans have a little more subjectivity in their hate speech laws. They may notify us about something that violates those laws, and if it does not violate our global standards, we’ll remove it in Germany only. (It’s done by IP addresses, the language you’re using, etc.) When we do that, we include it in our 6 month reports. If it’s removed, you see a notice that the content is restricted in your jurisdiction.


Q: Have you spoken to users about people from different cultures and backgrounds reviewing their content?

A: It’s a legitimate question. E.g., when it comes to nudity, even a room of people as homogenous as this one will disagree. So, “our rules are written to be very objective”our rules are written to be very objective. And we’re increasingly using tech to make these decisions. E.g., it’s easy to automate the finding of links to porn or spam, and much harder for evaluating speech.

Q: What drives change in these policies and algorithms?

A: It’s constantly happening. And public conversation is helpful. And our reviewers raise issues.

Q: a) When there are very contentious political issues, how do you prevent bias? b) Are there checks on FB promoting some agenda?

A: a) We don’t have a rule saying that people from one or another country can review contentious posts. But we review the reviewers’ decisions every week. b) The transparency report we put out every six months is one such check. If we don’t listen to feedback, we tend to see news stories calling us out on it.

[Monika now quickly addresses some of the questions from the open question tool.]

Q: Would you send reports to Lumen? MB: We don’t currently record why decisions were made.

Q: How to prevent removal policies from being weaponized but trolls or censorious regimes? MB: We treat all reports the same — there’s an argument that we shouldn’t — but we don’t continuously re-review posts.

JZ: For all of the major platforms struggling with these issues, is it your instinct that it’s just a matter of incrementally getting this right, bringing in more people, continue to use AI, etc. OR do you think sometimes that this is just nuts; there’s got to be a better way.

There’s a tension between letting anyone see what they want, or have global standards. People say US hates hate speech and the Germans not so much, but there’s actually a spectrum in each. The catch is that there’s content that you’re going to be ok seeing but we think is not ok to be shared.

[Monika was refreshingly direct, and these are, I believe, literally impossible problems. But I came away thinking that FB’s position has a lot to do with covering their butt at the expense of protecting the vulnerable. E.g., they treat all protected classes equally, even though some of us — er, me — are in top o’ the heap, privileged classes. The result is that FB applies a rule equally to all, which can bring inequitable results. That’s easier and safer, but it’s not like I have a solution to these intractable problems.]

1 Comment »

September 3, 2017

Free e-book from Los Angeles Review of Books

I’m proud that my essay about online knowledge has been included in a free e-book collecting essays about the effect of the digital revolution, published by the Los Angeles Review of Books.

It’s actually the first essay in the book, which obviously is not arranged in order of preference, but probably means at least the editors didn’t hate it.


The next day: Thanks to a tweet by Siva Vaidhyanathan, I and a lot of people on Twitter have realized that all but one of the authors in this volume are male. I’d simply said yes to the editors’ request to re-publish my article. It didn’t occur to me to ask to see the rest of the roster even though this is an issue I care about deeply. LARB seems to feature diverse writers overall, but apparently not so much in tech.

On the positive, this has produced a crowd-sourced list of non-male writers and thinkers about tech with a rapidity that is evidence of the pain and importance of this issue.

Comments Off on Free e-book from Los Angeles Review of Books

August 18, 2017

Journalism, mistrust, transparency

Ethan Zuckerman brilliantly frames the public’s distrust of institutional journal in a whitepaper he is writing for Knight. (He’s posted it both on his blog and at Medium. Choose wisely.)
As he said at an Aspen event where he led a discussion of it:

…I think mistrust in civic institutions is much broader than mistrust in the press. Because mistrust is broad-based, press-centric solutions to mistrust are likely to fail. This is a broad civic problem, not a problem of fake news,

The whitepaper explores the roots of that broad civic problem and suggests ways to ameliorate it. The essay is deeply thought, carefully laid out, and vividly expressed. It is, in short, peak Ethanz.

The best news is that Ethan notes that he’s writing a book on civic mistrust.



In the early 2000’s, some of us thought that journalists would blog and we would thereby get to know who they are and what they value. This would help transparency become the new objectivity. Blogging has not become the norm for reporters, although it does occur. But it turns out that Twitter is doing that transparency job for us. Jake Tapper (@jaketapper) at CNN is one particularly good example of this; he tweets with a fierce decency. Margie Haberman (@maggieNYT) and Glenn Thrush (@glennThrush) from the NY Times, too. And many more.

This, I think is a good thing. For one thing, it increases trust in at least some news media, while confirming our distrust of news media we already didn’t trust. But we are well past the point where we are ever going to trust the news media as a generalization. The challenge is to build public trust in news media that report as truthfully and fairly as they can.

1 Comment »

August 13, 2017

Machine learning cocktails

Inspired by fabulously wrong paint colors that Janelle Shane’s generated by running existing paint names through a machine learning system, and then by an hilarious experiment in dog breed names by my friend Matthew Battles, I decided to run some data through a beginner’s machine learning algorithm by karpathy.

I fed a list of cocktail names in as data to an unaltered copy of karpathy’s code. After several hundred thousand iterations, here’s a highly curated list of results:

  • French Connerini Mot
  • Freside
  • Rumibiipl
  • Freacher
  • Agtaitane
  • Black Silraian
  • Brack Rickwitr
  • Hang
  • boonihat
  • Tuxon
  • Bachutta B
  • My Faira
  • Blamaker
  • Salila and Tonic
  • Tequila Sou
  • Iriblon
  • Saradise
  • Ponch
  • Deiver
  • Plaltsica
  • Bounchat
  • Loner
  • Hullow
  • Keviy Corpse der
  • KreckFlirch 75
  • Favoyaloo
  • Black Ruskey
  • Avigorrer
  • Anian
  • Par’sHance
  • Salise
  • Tequila slondy
  • Corpee Appant
  • Coo Bogonhee
  • Coakey Cacarvib
  • Srizzd
  • Black Rosih
  • Cacalirr
  • Falay Mund
  • Frize
  • Rabgel
  • FomnFee After
  • Pegur
  • Missoadi Mangoy Rpey Cockty e
  • Banilatco
  • Zortenkare
  • Riscaporoc
  • Gin Choler Lady or Delilah
  • Bobbianch 75
  • Kir Roy Marnin Puter
  • Freake
  • Biaktee
  • Coske Slommer Roy Dog
  • Mo Kockey
  • Sane
  • Briney
  • Bubpeinker
  • Rustin Fington Lang T
  • Kiand Tea
  • Malmooo
  • Batidmi m
  • Pint Julep
  • Funktterchem
  • Gindy
  • Mod Brandy
  • Kkertina Blundy Coler Lady
  • Blue Lago’sil
  • Mnakesono Make
  • gizzle
  • Whimleez
  • Brand Corp Mook
  • Nixonkey
  • Plirrini
  • Oo Cog
  • Bloee Pluse
  • Kremlin Colone Pank
  • Slirroyane Hook
  • Lime Rim Swizzle
  • Ropsinianere
  • Blandy
  • Flinge
  • Daago
  • Tuefdequila Slandy
  • Stindy
  • Fizzy Mpllveloos
  • Bangelle Conkerish
  • Bnoo Bule Carge Rockai Ma
  • Biange Tupilang Volcano
  • Fluffy Crica
  • Frorc
  • Orandy Sour
  • The candy Dargr
  • SrackCande
  • The Kake
  • Brandy Monkliver
  • Jack Russian
  • Prince of Walo Moskeras
  • El Toro Loco Patyhoon
  • Rob Womb
  • Tom and Jurr Bumb
  • She Whescakawmbo Woake
  • Gidcapore Sling
  • Mys-Tal Conkey
  • Bocooman Irion anlis
  • Ange Cocktaipopa
  • Sex Roy
  • Ruby Dunch
  • Tergea Cacarino burp Komb
  • Ringadot
  • Manhatter
  • Bloo Wommer
  • Kremlin Lani Lady
  • Negronee Lince
  • Peady-Panky on the Beach

Then I added to the original list of cocktails a list of Western philosophers. After about 1.4 million iterations, here’s a curated list:

  • Wotticolus
  • Lobquidibet
  • Mores of Cunge
  • Ruck Velvet
  • Moscow Muáred
  • Elngexetas of Nissone
  • Johkey Bull
  • Zoo Haul
  • Paredo-fleKrpol
  • Whithetery Bacady Mallan
  • Greekeizer
  • Frellinki
  • Made orass
  • Wellis Cocota
  • Giued Cackey-Glaxion
  • Mary Slire
  • Robon Moot
  • Cock Vullon Dases
  • Loscorins of Velayzer
  • Adg Cock Volly
  • Flamanglavere Manettani
  • J.N. tust
  • Groscho Rob
  • Killiam of Orin
  • Fenck Viele Jeapl
  • Gin and Shittenteisg Bura
  • buzdinkor de Mar
  • J. Apinemberidera
  • Nickey Bull
  • Fishomiunr Slmester
  • Chimio de Cuckble Golley
  • Zoo b Revey Wiickes
  • P.O. Hewllan o
  • Hlack Rossey
  • Coolle Wilerbus
  • Paipirista Vico
  • Sadebuss of Nissone
  • Sexoo
  • Parodabo Blazmeg
  • Framidozshat
  • Almiud Iquineme
  • P.D. Sullarmus
  • Baamble Nogrsan
  • G.W.J. . Malley
  • Aphith Cart
  • C.G. Oudy Martine ram
  • Flickani
  • Postine Bland
  • Purch
  • Caul Potkey
  • J.O. de la Matha
  • Porel
  • Flickhaitey Colle
  • Bumbat
  • Mimonxo
  • Zozky Old the Sevila
  • Marenide Momben Coust Bomb
  • Barask’s Spacos Sasttin
  • Th mlug
  • Bloolllamand Royes
  • Hackey Sair
  • Nick Russonack
  • Fipple buck
  • G.W.F. Heer Lach Kemlse Male

Yes, we need not worry about human bartenders, cocktail designers, or philosophers being replaced by this particular algorithm. On the other hand, this is algorithm consists of a handful of lines of code and was applied blindly by a person dumber than it. Presumably SkyNet — or the next version of Microsoft Clippy — will be significantly more sophisticated than that.

Comments Off on Machine learning cocktails

August 8, 2017

Messy meaning

Steve Thomas [twitter: @stevelibrarian] of the Circulating Ideas podcast interviews me about the messiness of meaning, library innovation, and educating against fake news.

You can listen to it here.

Comments Off on Messy meaning

« Previous Page | Next Page »