Causes of (and Remedies for) Bias in AI

James Manyika, Jake Silberg, and Brittany Presten writing for the Harvard Business Review:

AI can help identify and reduce the impact of human biases, but it can also make the problem worse by baking in and deploying biases at scale in sensitive application areas.

The phrase “artificial intelligence” is leading us astray. For some folks, it’s become a type of magical incantation that promises to solve all sorts of problems. Much of what goes by AI today isn’t magic — or intelligence, really; it’s dynamic applied statistics. As such, “AI” is highly subject to the data being analyzed and the structure of that data. Garbage in, garbage out.

It’s important for business leaders to learn about how AI works. The HBR post offers a good summary of the issues and practical recommendations for leaders looking to make better decisions when implementing AI-informed systems — which we all should be:

Bias is all of our responsibility. It hurts those discriminated against, of course, and it also hurts everyone by reducing people’s ability to participate in the economy and society. It reduces the potential of AI for business and society by encouraging mistrust and producing distorted results. Business and organizational leaders need to ensure that the AI systems they use improve on human decision-making, and they have a responsibility to encourage progress on research and standards that will reduce bias in AI.

What Do We Do About the Biases in AI?

Being Open to Unsettling Changes

My post about watching movies at double-speed elicited strong reactions. Some folks seem convinced that giving people the ability to watch movies faster will diminish the viewing experience, and not just for them — for everyone. Why? Because such changes inevitably influence the medium itself.

Consider the way sound changed movies. “Talking” pictures did away with title cards. That was a significant change to the medium, which was wrought by advances in technology. Once it was possible to synchronize sound and pictures, irreversible changes to the medium were inevitable.

Are movies with sound better or worse than what came before? That’s a judgment call. It depends on your point of view. You and I grew up in a world of talking pictures; the silent ones with their title cards seem old and clunky. But they have merits too. Silent films had literate value. Many featured live musical performances, which made them into more of an event than pre-recorded movies. I can imagine somebody who grew up with silent movies could become attached to the way they were.

Continue reading

Binge Faster

According to a report on The Verge, Netflix is testing a way for its users to watch shows and movies at double-speed. Some people associated with the movie industry are pushing back. The report includes a quote from director Brad Bird that captures the sentiment:

Why support and finance filmmakers visions on one hand and then work to destroy the presentation of those films on the other?

“Destroy the presentation” sounds like an exaggeration. But in cinema, timing is critical. Directors, actors, and editors obsess over getting the pacing of scenes and dialog just right. Giving users the ability to speed everything up can ruin the intended effect. So why would Netflix do this? Per the report, “it’s a heavily requested feature from subscribers.”

I’m not surprised. In our era of binge-watching (driven in part by Netflix) and shortening attention spans, speeding up shows would allow users to watch more. I can relate to the sentiment. I listen to a lot of audiobooks, and often I’ll do so at 2x or sometimes even 2.5x the regular speed. As a result, I can read more in less time.

Not all books work well when sped up; some I can only listen to at 1x. Mostly, the “slow” ones are books I want to savor — either because the story is gripping, or I’m enjoying the narrator’s performance. Case in point: I’m currently listening to the Sherlock Holmes stories read by Stephen Fry. The audiobook is almost 63 hours long, and I’ve read all of it at 1x; anything else would ruin Mr. Fry’s terrific performance.

Books that work well when sped up are the ones I read “for work” instead of “for pleasure.” (In quotes because this isn’t a hard distinction for me.) The main challenge with these is taking notes whenever something interesting comes up. This is harder to do when listening fast. But in these cases, I’m looking for information, not performance. So faster speeds work for me.

Now, you could argue that people watch shows on Netflix (and other streaming services) mostly “for pleasure.” But why not leave it up to them to determine how they want to listen? It’s not like everyone would be forced to watch at higher speeds; it’s just a new choice.

Netflix wants to let people watch things at twice the speed, but Hollywood is pushing pack

Quantum Supremacy

Earlier this week, Google researchers announced a major computing breakthrough in the journal Nature:

Our Sycamore processor takes about 200 seconds to sample one instance of a quantum circuit a million times—our benchmarks currently indicate that the equivalent task for a state-of-the-art classical supercomputer would take approximately 10,000 years. This dramatic increase in speed compared to all known classical algorithms is an experimental realization of quantum supremacy for this specific computational task, heralding a much-anticipated computing paradigm.

Quantum supremacy heralds an era of not merely faster computing, but one in which computers can solve new types of problems. There was a time when I’d expect such breakthroughs from “information technology” companies such as IBM. But Google’s tech is ultimately in service to another business: advertising.

TAOI: Facebook’s Electoral Interference Changes

The architecture of information:

The Verge reports on a set of changes to Facebook aimed at counteracting misinformation on its platform. The changes come ahead of next years’ elections in the U.S., and include tools to protect candidates’ accounts, more transparency about the entities that manage Facebook pages, and new advertiser guidelines.

Reading through this list reminds me of the role television has played in influencing electoral outcomes. Compared to an information environment like Facebook, television — even in its current state, with hundreds of channels to choose from — has limited bandwidth. As a result, both actors and gatekeepers must be selective about what they publish on TV.

Compared to television, publishing on digital social platforms is cheap and easy. Anyone can publish anything, including variations on ads so they can be optimized for effectiveness. Additionally, on a social platform like Facebook, the people who are being influenced can also be publishers — that is, they can help spread messages “virally.” As a result, digital is a more effective platform for persuasion than TV.

I’m glad to see Facebook making structural changes to increase transparency and trustworthiness of their platform. Given its scale and reach, these changes could have an impact on the fairness of elections.

Facebook will label false posts more clearly as part of an effort to prevent 2020 election interference

Paola Antonelli on the Role of Designers

MoMA Senior Curator of Architecture and Design Paola Antonelli in an interview with Core77:

In 2008, I did an exhibition that was called Design and the Elastic Mind where I came up with this idea that I use a lot: that designers are enzymes, they are the ones that make innovation, whether it’s scientific or technological into life. So that’s how I always think about it, whether they’re working on a product or on an interface or on an exhibition design, designers are the ones who make sure that there’s a synthesis happening and the synthesis that can be communicated to other human beings.

So if they are exhibition designers they make sure that the idea of a curator is tangible and is understandable by other people. If they are product designers, they in the simplest of cases become an interface between the engineering department and the public.

So it really depends, but I believe the designers are naturally extroverted professionals. Not that they are individually extroverted, maybe not, they might be shy or mostly introverts. But their role is to become catalysts, enzymes and to put the pieces together, they’re very good at that. Their role in the future continues to be that and I hope that this exquisite characteristic they have will be used and exploited in areas that are not necessarily the usual one. I hope that they will be included in political discussions, I hope that they will be almost like philosophers who are society-wise people that are consulted whenever there’s a big decision to make.

Design isn’t just a way of making better products and services, it’s also a way of knowing the world. (I love that Ms. Antonelli extends the scope of design to include politics, which has more to do with tweaking social and economic relations than with making stuff.)

Paola Antonelli on the Urgent Role Designers Will Play in the Future

Perspective on “Digital”

Twenty-five years ago, I left my career in architecture. I’d been working in the building-design trade for about a year. Then something happened that led me to abandon the profession I’d trained for and which I’d pined for only a few years earlier: I saw Mosaic, the first widely available web browser.

I’d been aware of the Internet while I was a university student. I thought of it as a command-line system, one that was mostly useful for email. But the web was different. It was visual. It easy to use, both for end-users and content authors. Anyone could publish at global scale, cheaply. It was clear to me that making the world’s information available through the web would change our reality in significant ways. I cast my lot with “digital” and haven’t looked back.

A quarter of a century later, I still love the web. I love what it’s become – even with its (many) flaws. But more importantly, I love what it can be — its latent potential to improve the human condition. There’s so much work to be done.

But there’s a lot of negativity centered on digital technology these days. Here’s a sample of headlines from major publications from the last few months:

These stories are representative of a melancholic tone that’s become all too common these days. Our pervasive digital technologies have wrought significant changes in old ways of being in the world. Some people miss the old ways; others are perplexed or alarmed. It’s understandable, but it couldn’t be otherwise. The internet and the constellation of services it enables are profoundly disruptive.

Social structures don’t remain unchanged after encountering such forces. The introduction of the printing press led to social, political, and scientific revolutions — including the Reformation. These weren’t small, incremental changes to the social fabric; they shattered then-current reality and re-configured it in new and surprising ways. The process wasn’t smooth or easy.

Digital is more radically transformative than the printing press. It’s folly to expect long-established social structures will stand as before. The question isn’t whether our societies will change, but whether the change will be for better or worse. That’s still an open question. I’m driven by the idea that those of us working in digital have the opportunity to help steer the outcome towards humane ends.

TAOI: Facebook Hiding Likes

The Architecture of Information:

Likes are one of the most important concepts of the Facebook experience. Giving users the ability to cast their approval (or disapproval) on a post or comment — and to see how others have “voted” — is one of the most engaging aspects of the system, both for users and content authors. Facebook even uses the Like icon as a symbol of the company as a whole:

fbwm_cw_07
The sign outside the main entrance to Facebook headquarters. (Photo: Facebook.)

However, according to a report in the NY Times, Facebook is experimenting with hiding post measurements:

On [September 26], the social network said it was starting a test in Australia, where people’s Likes, video view counts and other measurements of posts would become private to other users. It is the first time the company has announced plans to hide the numbers on its platform.

Why would they do this? Because seeing these metrics may have an impact on users’ self-esteem. According to a Facebook spokesperson quoted in the article, the company will be testing the change to see if it helps improve people’s experiences. A noble pursuit. But, I wonder: How would this impact user engagement? If it benefits users but hurts advertising revenue, will Facebook discontinue the experiment?

Facebook Tests Hiding ‘Likes’ on Social Media Posts

Collaborating by Default

Writing in his blog, Benedict Evans highlights the new wave of startups focused on personal productivity, “dozens of companies that remix some combination of lists, tables, charts, tasks, notes, light-weight databases, forms, and some kind of collaboration, chat or information-sharing.”

The cycle of bundling and unbundling functionality isn’t new:

There’s an old joke that every Unix function became an internet company – now every Craigslist section, or LinkedIn category, or Excel template, becomes a company as well. Depending on the problem, that might be a new collaboration canvas, or a new networked app, or a new network or marketplace, and you might switch from one form to the other. Github is a developer tool that also became a network – it became LinkedIn for developers.

What is new is the social nature of the experience. Old-school computing was lonely: the user interacted with his/her computer alone. Even if the system included communications software, such as email, interactions with other people were limited to that software alone. Today, we expect web-based applications to be collaborative by default.

We experience software differently when we assume other people will be sharing the place with us. As I’ve written before, we may ultimately discover that the purpose of social media was to teach us how to collaborate with people in information environments.

New Productivity