Connect with us

Uncategorized

Tips for applying an intersectional framework to AI development

Published

on

By now, most of us in tech know that the inherent bias we possess as humans creates an inherent bias in AI applications — applications that have become so sophisticated they’re able to shape the nature of our everyday lives and even influence our decision-making.

The more prevalent and powerful AI systems become, the sooner the industry must address questions like: What can we do to move away from using AI/ML models that demonstrate unfair bias?

How can we apply an intersectional framework to build AI for all people, knowing that different individuals are affected by and interact with AI in different ways based on the converging identities they hold?

Start with identifying the variety of voices that will interact with your model.

Intersectionality: What it means and why it matters

Before tackling the tough questions, it’s important to take a step back and define “intersectionality.” A term defined by Kimberlé Crenshaw, it’s a framework that empowers us to consider how someone’s distinct identities come together and shape the ways in which they experience and are perceived in the world.

This includes the resulting biases and privileges that are associated with each distinct identity. Many of us may hold more than one marginalized identity and, as a result, we’re familiar with the compounding effect that occurs when these identities are layered on top of one another.

At The Trevor Project, the world’s largest suicide prevention and crisis intervention organization for LGBTQ youth, our chief mission is to provide support to each and every LGBTQ young person who needs it, and we know that those who are transgender and nonbinary and/or Black, Indigenous, and people of color face unique stressors and challenges.

So, when our tech team set out to develop AI to serve and exist within this diverse community — namely to better assess suicide risk and deliver a consistently high quality of care — we had to be conscious of avoiding outcomes that would reinforce existing barriers to mental health resources like a lack of cultural competency or unfair biases like assuming someone’s gender based on the contact information presented.

Though our organization serves a particularly diverse population, underlying biases can exist in any context and negatively impact any group of people. As a result, all tech teams can and should aspire to build fair, intersectional AI models, because intersectionality is the key to fostering inclusive communities and building tools that serve people from all backgrounds more effectively.

Doing so starts with identifying the variety of voices that will interact with your model, in addition to the groups for which these various identities overlap. Defining the opportunity you’re solving is the first step because once you understand who is impacted by the problem, you can identify a solution. Next, map the end-to-end experience journey to learn the points where these people interact with the model. From there, there are strategies every organization, startup and enterprise can apply to weave intersectionality into every phase of AI development — from training to evaluation to feedback.

Datasets and training

The quality of a model’s output relies on the data on which it’s trained. Datasets can contain inherent bias due to the nature of their collection, measurement and annotation — all of which are rooted in human decision-making. For example, a 2019 study found that a healthcare risk-prediction algorithm demonstrated racial bias because it relied on a faulty dataset for determining need. As a result, eligible Black patients received lower risk scores in comparison to white patients, ultimately making them less likely to be selected for high-risk care management.

Fair systems are built by training a model on datasets that reflect the people who will be interacting with the model. It also means recognizing where there are gaps in your data for people who may be underserved. However, there’s a larger conversation to be had about the overall lack of data representing marginalized people — it’s a systemic problem that must be addressed as such, because sparsity of data can obscure both whether systems are fair and whether the needs of underrepresented groups are being met.

To start analyzing this for your organization, consider the size and source of your data to identify what biases, skews or mistakes are built-in and how the data can be improved going forward.

The problem of bias in datasets can also be addressed by amplifying or boosting specific intersectional data inputs, as your organization defines it. Doing this early on will inform your model’s training formula and help your system stay as objective as possible — otherwise, your training formula may be unintentionally optimized to produce irrelevant results.

At The Trevor Project, we may need to amplify signals from demographics that we know disproportionately find it hard to access mental health services, or for demographics that have small sample sizes of data compared to other groups. Without this crucial step, our model could produce outcomes irrelevant to our users.

Evaluation

Model evaluation is an ongoing process that helps organizations respond to ever-changing environments. Evaluating fairness began with looking at a single dimension — like race or gender or ethnicity. The next step for the tech industry is figuring out how to best compare intersectional groupings to evaluate fairness across all identities.

To measure fairness, try defining intersectional groups that could be at a disadvantage and the ones that may have an advantage, and then examine whether certain metrics (for example, false-negative rates) vary among them. What do these inconsistencies tell you? How else can you further examine which groups are underrepresented in a system and why? These are the kinds of questions to ask at this phase of development.

Developing and monitoring a model based on the demographics it serves from the start is the best way for organizations to achieve fairness and alleviate unfair bias. Based on the evaluation outcome, a next step might be to purposefully overserve statistically underrepresented groups to facilitate training a model that minimizes unfair bias. Since algorithms can lack impartiality due to societal conditions, designing for fairness from the outset helps ensure equal treatment of all groups of individuals.

Feedback and collaboration

Teams should also have a diverse group of people involved in developing and reviewing AI products — people who are diverse not only in identities, but also in skillset, exposure to the product, years of experience and more. Consult stakeholders and those who are impacted by the system for identifying problems and biases.

Lean on engineers when brainstorming solutions. For defining intersectional groupings, at The Trevor Project, we worked across the teams closest to our crisis-intervention programs and the people using them — like Research, Crisis Services and Technology. And reach back out to stakeholders and people interacting with the system to collect feedback upon launch.

Ultimately, there isn’t a “one-size-fits-all” approach to building intersectional AI. At The Trevor Project, our team has outlined a methodology based on what we do, what we know today and the specific communities we serve. This is not a static approach and we remain open to evolving as we learn more. While other organizations may take a different approach to build intersectional AI, we all have a moral responsibility to construct fairer AI systems, because AI has the power to highlight — and worse, magnify — the unfair biases that exist in society.

Depending on the use case and community in which an AI system exists, the magnification of certain biases can result in detrimental outcomes for groups of people who may already face marginalization. At the same time, AI also has the ability to improve quality of life for all people when developed through an intersectional framework. At The Trevor Project, we strongly encourage tech teams, domain experts and decision-makers to think deeply about codifying a set of guiding principles to initiate industry-wide change — and to ensure future AI models reflect the communities they serve.

Lyron Foster is a Hawaii based African American Musician, Author, Actor, Blogger, Filmmaker, Philanthropist and Multinational Serial Tech Entrepreneur.

Continue Reading
Comments

Uncategorized

Tim Hortons marks two years in China with Tencent investment

Published

on

Tim Hortons, the Canadian coffee and doughnut giant, has raised a new round of funding for its Chinese venture. The investment is led by Sequoia China with participation from Tencent, its digital partner in China, and Eastern Bell Capital. The round comes two years after Tim Hortons made its foray into China’s booming coffee industry.

Tim Hortons didn’t disclose the amount of its latest fundraise but noted in a social media post that the proceeds will be used for opening more stores, building its digital infrastructure, brand presence, and more.

Tencent, the Chinese social media and entertainment behemoth, first backed the 57-year-old Canadian coffee chain last May. At the time the tie-up was seen as Tencent’s move to counter archrival Alibaba’s alliance with Starbucks to deliver coffee and help the American coffee titan go digital in China.

Tim Horton’s collaboration with the WeChat parent is in a similar vein. It has so far accumulated three million members through its WeChat mini program, a type of lightweight app that runs within the instant messenger. To appeal to young Chinese consumers, Tim Hortons opened an esports-themed cafe with Tencent, China’s biggest gaming company.

Two years into operating in China, Tim Hortons says it has reached storefront-level profitability with a footprint of 150 locations across 10 major cities. It plans to add more than 200 locations in 2021 and reach 1,500 stores nationwide in the next few years.

The dramatic rise and fall of coffee delivery startup Luckin brought the prospects of China’s coffee market to the forefront. Despite the investment frenzy around Luckin and other coffee businesses, coffee drinking still has a relatively low penetration in China compared to countries like the United States and Germany. On the other hand, coffee consumption is growing at a much faster rate of 15% in China, well above the global average of 2%, and is projected to reach 1 trillion yuan ($150 million) in 2025, according to a 2020 report by Dongxing Securities.

Continue Reading

Uncategorized

Bessemer Venture Partners closes on $3.3 billion across two funds

Published

on

Another major VC firm has closed two major rounds, underscoring the long-term confidence investors continue to have for backing privately-held companies in the tech sector.

Early-stage VC firm Bessemer Venture Partners announced Thursday the close of two new funds totaling $3.3 billion that it will be using both to back early-stage startups as well as growth rounds for more mature companies.

The Redwood City-based firm closed BVP XI with $2.475 billion and BVP Century II with $825 million in total commitments.

With BVP XI, it plans to focus on early-stage companies spanning across enterprise, consumer, healthcare, and frontier technologies. 

Its Century II fund is aimed at backing growth-stage companies that Bessemer believes “will define the next century,” and will include both follow-on rounds for existing portfolio companies or investments in new ones.

BVP XI marks Bessemer’s largest fund in its 110-year history. In October 2018, the firm brought in $1.85 billion for its tenth flagship VC fund. This latest fund is its fifth consecutive billion-dollar fund, based on PitchBook data. 

Despite being founded more than 100 years ago, Bessemer didn’t actually enter the venture business until 1965. It’s known for its investments in LinkedIn, Blue Apron and many others, with a current portfolio that includes PagerDuty, Shippo, Electric and DocuSign. Exits include Twitch and Shopify, among many others.

With more money than ever before available for backing startups, the challenge now for VCs is to see how and if they can find (and invest in) whatever will define the next generation of tech. 

“As venture capitalists, we pay too much attention to pattern recognition and matching when in reality, the biggest opportunities exist where those patterns break,” the firm wrote in a blog post today. “Our job is to make perceptive bets on the future, especially those that others will dismiss and ridicule. We are fundamental optimists and strong believers in the power of innovation; our life’s work is putting our reputation, time, and money to help entrepreneurs realize a different future. They’re the ones pioneering something entirely new and obscure – a technology, a business model, a category.

In addition to announcing the new funds, Bessemer also revealed today that it’s brought on five new partners including Jeff Blackburn, who joins after a 22-year career at Amazon, alongside the promotion of existing investors Mary D’Onofrio, Mike Droesch, Tess Hatch, and Andrew Hedin.

Most recently at Amazon, Blackburn served as senior vice president of worldwide business development where he oversaw dozens of Amazon’s minority investments and more than 100 acquisitions across all business lines – including retail, Kindle, Echo, Alexa, FireTV, advertising, music, streaming audio & video, and Amazon Web Services.  

“Having been part of Amazon for more than two decades, I’m excited to begin a new chapter helping customer-focused founders build breakthrough companies,” said Blackburn in a written statement.  “I’ve known the Bessemer team for many years and have long admired their strategic vision and success backing early-stage ventures.” 

With the latest changes, Bessemer now has 21 partners and over 45 investors, advisors, and platform “team members” located in Silicon Valley, San Francisco, Seattle, New York, Boston, London, Tel Aviv, Bangalore, and Beijing. 

“At Bessemer, there’s no corner office or consensus; every partner has the choice, independently, to pen a check. This kind of accountability and autonomy means a founder is teaming up with a partner and board director who thoroughly understands your business and can respond quickly and decisively,” the firm’s blog post read.

Continue Reading

Uncategorized

Daily Crunch: Twitter announces ‘Super Follow’ subscriptions

Published

on

Twitter reveals its move into paid subscriptions, Australia passes its media bargaining law and Coinbase files its S-1. This is your Daily Crunch for February 25, 2021.

The big story: Twitter announces ‘Super Follow’ subscriptions

Twitter announced its first paid product at an investor event today, showing off screenshots of a feature that will allow users to subscribe to their favorite creators in exchange for things like exclusive content, subscriber-only newsletters and a supporter badge.

The company also announced a feature called Communities, which could compete with Facebook Groups and enable Super Follow networks to interact, plus a Safety Mode for auto-blocking and muting abusive accounts. On top of all that, Twitter said it plans to double revenue by 2023.

Not announced: launch dates for any of these features.

The tech giants

After Facebook’s news flex, Australia passes bargaining code for platforms and publishers — This requires platform giants like Facebook and Google to negotiate to remunerate local news publishers for their content.

New Facebook ad campaign extols the benefits of personalized ads — The sentiments are similar to a campaign that Facebook launched last year in opposition to Apple’s upcoming App Tracking Transparency feature.

Startups, funding and venture capital

Sergey Brin’s airship aims to use world’s biggest mobile hydrogen fuel cell — The Google co-founder’s secretive airship company LTA Research and Exploration is planning to power a huge disaster relief airship with an equally record-breaking hydrogen fuel cell.

Coinbase files to go public in a key listing for the cryptocurrency category — Coinbase’s financials show a company that grew rapidly from 2019 to 2020 while also crossing the threshold into unadjusted profitability.

Boosted by the pandemic, meeting transcription service Otter.ai raises $50M — With convenient timing, Otter.ai added Zoom integration back in April 2020.

Advice and analysis from Extra Crunch

DigitalOcean’s IPO filing shows a two-class cloud market — The company intends to list on the New York Stock Exchange under the ticker symbol “DOCN.”

Pilot CEO Waseem Daher tears down his company’s $60M Series C pitch deck — For founders aiming to entice investors, the pitch deck remains the best way to communicate their startup’s progress and potential.

Five takeaways from Coinbase’s S-1 — We dig into Coinbase’s user numbers, its asset mix, its growing subscription incomes, its competitive landscape and who owns what in the company.

(Extra Crunch is our membership program, which helps founders and startup teams get ahead. You can sign up here.)

Everything else

Paramount+ will cost $4.99 per month with ads — The new streaming service launches on March 4.

Register for TC Sessions: Justice for a conversation on diversity, equity and inclusion in the startup world — This is just one week away!

The Daily Crunch is TechCrunch’s roundup of our biggest and most important stories. If you’d like to get this delivered to your inbox every day at around 3pm Pacific, you can subscribe here.

Continue Reading

Trending