When passion and technology meet

Ash Bhat and Rohan Phadte have been friends since junior high in San Jose, California, when Bhat dove into coding and Phadte tinkered with robots.

Now, they are barely past 20 years old, housemates in Berkeley, California, the headquarters for their startup RoBhat Labs, fighting the scourges of our digital political reality.

In May, the men used some of the methods they had picked up in machine learning classes to help develop a NewsBot for Facebook Messenger: Drop a link of a news article into the chat window and the bot judges whether the report has right or left bias. Last summer, they started working on an algorithm to identify the bots on Twitter pushing inflamed political discussion.

RoBhat Labs introduced Botcheck.me, a browser extension that lets Twitter users test whether a Twitter account is run by a human or is automated. The associated website also shows what bot networks are tweeting about in real time. The Democratic National Committee took notice, commissioning RoBhat Labs to write a report on the spread of disinformation on social media.

“I’m excited about them stepping up and applying public pressure” on the tech companies, said Raffi Krikorian, the committee’s chief technology officer. “We frankly think the administration and Congress are not doing enough, so we’re filling in this gap.”

Bhat and Phadte left school this spring to work with the committee and are seeking funding for their startup. They’re on a deadline to develop their next product, which will be a tool that tests whether news has been altered as it gets passed around the web, warding against doctored photos and altered videos. They hope to market the tool to news media companies and political candidates this summer, in time for the run-up to the midterms.

______

Dealing With Crises in a Text-Savvy Generation

Nancy Lublin

Chief executive of Crisis Text Line, crisistextline.org

When Nancy Lublin started Crisis Text Line in 2013, she thought of her New York City nonprofit as a tech startup — and not simply because they text to counsel people in distress. Lublin wanted to use analytics to be smarter at it.

“I’d consider myself a tech CEO,” she said. “The first thing I ask is, ‘What does the data say?'”

Five years after the site began, 4,000 volunteer counselors across the country — the site is always recruiting more — can work on their computers wherever there is internet. The nonprofit said it handled nearly 1 million chats last year, mostly with people under 25 who would rather text than talk about their problems.

The platform the volunteers log into is honed with a machine learning algorithm, which has analyzed past chats for actionable information. For example, the system automatically moves distressed people who write words like “gun,” “military,” “fentanyl,” or a crying emoji to the front of the line, colored orange. Data shows those are the most likely indicators that the person will need an active rescue from 911, and counselors start chats with them first, responding within an average of 38 seconds.

The chats are also creating a trove of information on distress nationwide. The highest per capita texts come from Montana; Midwesterners report the most bullying; the coasts are the most stressed; and people in the South most often mention LGBTQ issues. (You can track the state-level data in real time at crisistrends.org.) The organization’s tech team is working on software to make the counselors even faster: for example, suggesting they answer with “you’re strong” to people who say they are overwhelmed.

The data also helps people on the ground who are fighting crises. In 2016, Ohio started a state-specific keyword with the Crisis Text Line — 4HOPE — and advertised it in schools. The nonprofit’s data crunchers noticed that use of the keyword fell in the summer. Their Ohio partners were alerted and placed advertising in movie theaters. The texts surged once more.

______

From Soup to Brooms, Changing the World

Eileen Richardson

Chief executive of Downtown Streets Team, streetsteam.org

In 2005, Eileen Richardson started volunteering at a soup kitchen in Silicon Valley. She was chief executive of the music-file sharing platform Napster during the first tech boom, and she brought the same rigor to her volunteer position. She figured — in retrospect, with an embarrassing amount of hubris — that she could solve homelessness in six months and move on to her next project.

“That’s how tech people are,” she said. Instead, “Here I am all these years later. We’re onto something.”

Her solution is the Downtown Streets Team. Each weekday morning in nine California cities, team members sweep the streets for a four-hour shift. Led by other team members who have been promoted to managers, the daily shift teaches job skills and accountability, and boosts the team members’ self-esteem.

“You go from being a bad guy to now one of the great guys,” Richardson said of the members. “The chief of police and mayor are patting your back, and people are slipping you 20 bucks here and there.” The brigade of sweepers in yellow T-shirts acts as its own recruitment device: Other homeless people notice and ask to join.

In exchange for sweeping, team members receive stipends — gift cards for groceries, vouchers for housing, bus passes and prescription refills. They set goals with a case manager for housing and employment, and attend weekly team meetings that take on a nearly ecclesiastical fervor as everyone cheers each other’s successes. “This is not a program,” Richardson said. “It’s a team. They start seeing other people doing great and are like, ‘I want to get a car or housing, so it’s self motivating.'”

Thirteen years after Richardson’s experiment — with California in a housing crisis and a homeless epidemic — team members average four months to get housing and six months until they get a job. In a recent survey, 96 percent said they had more hope, 84 percent had health insurance and 73 percent said they were using less drugs and alcohol. Richardson is now looking to scale up with an affiliate program nationwide.

______

Crunching Numbers to Help People in India

Prukalpa Sankar and Varun Banka

Founders of SocialCops, socialcops.com

Five years ago, Prukalpa Sankar was studying business at a university in Singapore and interning at Goldman Sachs, the beginning of what she thought would be a decade of studying and working abroad before moving back to her native India. That all changed in a late-night dorm brainstorm, when she and another Indian classmate, Varun Banka, who was interning at Barclays, wondered why the same amount of big data they saw informing financial decisions was not being applied to the gnarlier problems in their home country. “We thought, wouldn’t it be so cool to use data to allocate where money gets spent in Indian cities?” Sankar said.

They raised some seed capital through business plan competitions ($25,000, enough to get started) and moved to New Delhi after graduation in 2013 to test a few pilots with a data science startup they named SocialCops.

They discovered that much of the data in the world’s second most populous country was still gathered by pencil and paper, in local languages, and never converted digitally — making any sort of big-picture analysis impossible. “It’s so difficult to reach data you can use, people often give up in the process,” Sankar, 26, said. “Yet when you can give people insights off the data, they will make decisions off of it.”

In 2016, India’s federal government enlisted their help to get propane tanks to women living in villages who had been using smoke-billowing indoor fires to cook. The government wanted to install gas cylinders in 50 million homes in three years and have public oil companies open thousands more centers nearby to sell them refills.

SocialCops went to work, asking 17,000 oil distributors to submit their GPS coordinates using the company’s app. They then incorporated data like populations, affluence and distance from existing sales centers, and made all the information easy to analyze to pinpoint the best locations for new centers and track the rollout of the connections in homes. The company said the effort hooked up about 22 million homes in its first year, 7 million more than the goal.

______

Delivering Messages to Aid Farmers in Ghana

Alloysius Attah and Emmanuel Owusu Addai

Co-founders of Farmerline, farmerline.co

Alloysius Attah grew up with his aunt, a farmer who grew corn, cassava and yams in rural Ghana. “She’s one of the hardest working and resilient people I’ve known,” he said. “I was a reluctant kid helping.” Attah said farmers were more than half the country’s labor force and many tended to 2 to 4 acres. Their access to information is limited, coming from local Agriculture Ministry offices or radio.

Attah, now 28, wants to give farmers the tools to become informed entrepreneurs, increasing their crop yields and marketability to the lucrative export business. Ghana is one of the world’s top exporters of cocoa, he said. “It’s a billions of dollar industry, but when you look at the money that comes down to farmers — they don’t get enough. That’s why we help them to tap into the global market.”

In 2013, he started Farmerline with his co-founder, Emmanuel Owusu Addai, which is based in Ghana and has grown into a company of 32 employees. Farmers nearly universally own basic cellphones, so Farmerline delivers them prerecorded voice messages in their local language with localized weather forecasts and market prices for crops. (“It’s like a podcast on a ‘dumb phone,'” Attah wrote in an email.)

The messages also share farming tips — like how to get certified as fair trade farmers, or increase crop yields by, say, pruning cacao trees to increase the growth of pods, or how to meet the crop specifications for export markets.

Farmerline uses machine learning to predict the local demand for farming supplies like fertilizer and equipment to then negotiate better prices for good quality goods from distributors. They also are using artificial intelligence to create a credit score, so farmers can take out loans to grow their business.

Overall, it puts farmers “in a position of power,” Attah said. “They move from a position of let me take what everybody gives me, to let me choose because I have options now. You and I have that.”

______

How We Picked Our Visionaries

People love lists.

We want to check out the best places to travel, catch up with the best inventions of the last 100 years, be in the know about the best-dressed people, the best books, the best schools. And on and on.

Of course, there is a risk to listmaking. Maybe your choices won’t hold up over the years. Maybe the best book of decades ago seems not so great today.

With the listmaking fervor and its risks in mind, we searched for people who would fit our criteria for visionaries. They had to be people who are forward-looking, working on exciting projects, helping others or taking a new direction. We wanted diversity in gender, race and ethnic background.

We assigned writers who are knowledgeable about the subjects we deemed most important. And we limited the list to 30.

Narrowing down the numbers was a huge challenge. And that’s a good problem to have. It means there are a lot of people out there who are following their visions.

We hope this inspires you to follow yours.

This article originally appeared in The New York Times.