Technology for Good: Author Jim Fruchterman in Conversation

[MUSIC PLAYING]

VOICEOVER: Technology for Good: Author Jim Fruchterman in Conversation. Speakers: Jim Fruchterman, Author and CEO, Tech Matters. Moderator: Larry Goldberg, Accessibility Sensei and Technology Consultant.

LARRY GOLDBERG: Hi, this is Larry Goldberg, uh, here with Jim Fruchterman. We wanna thank Sight Tech Global for giving us a few minutes to talk about some pretty interesting developments in the field of technology and accessibility and assistive tech. Hi, Jim.

JIM FRUCHTERMAN: Hey, Larry. Great to see you.

LARRY GOLDBERG: Great to see you. Now, before we dig in, I wanna talk about the most exciting startup I’ve heard of launch in the past year, and that is the Dovetail Bakery in Queens.

JIM FRUCHTERMAN: Yay. My daughter’s bakery.

LARRY GOLDBERG: Fantastic. And I’m looking out my window in New York right now and I could see Queens, so I’m gonna get over there.

JIM FRUCHTERMAN: Yeah. Yeah.

LARRY GOLDBERG: Congratulations. That’s great.

JIM FRUCHTERMAN: Thank you. You know, the better part of my family.

LARRY GOLDBERG: Yeah, absolutely. Now, I know you love to convent- uh, question conventional wisdom. So, you know, let’s dig into a few provocative things you brought up in your recent book. Um, your book is great. Just came out month and a half ago, two months ago. And it’s pretty exciting and people really need to read it. But why don’t you do a little promo for your book right now?

JIM FRUCHTERMAN: Well, obviously, you know, I dropped out of the regular Silicon Valley AI scene to start a nonprofit tech company that people know as Benetech. And, uh… And it was such a great experience for me to move away from, I call it, from money to meaning. And as my career has progressed, I have found hundreds of nonprofits that are using technology to make a difference, and no one had written a book on how to start a deliberately nonprofit tech company. Um, and so I thought there should be at least one book on how to deliberately lose money as a tech startup. (laughs)

LARRY GOLDBERG: And so you have. So you’re like… Well, my first question was, have you gotten any pushback from some of those nonprofits or even the dinosaurs that you talk about- (laughs) … that kind of need to go away?

JIM FRUCHTERMAN: You know, the dinosaurs aren’t reading my book, right? And, and, and, and the, and the joke in my, in my thing is like, you know, and, and the accessibility industry was kind of like… has had this issue, is that programs that were great in the 1950s maybe didn’t serve people so well in the 1990s or the 2010s, right? And yet, in the nonprofit sector, there’s a conservative… Conservative as in preserving the way we did things in the past approach and, and that tech innovation is often put in the back burner. I mean, the Library of Congress was shipping audio cassette tapes through the mail 25 years after the audio cassette tape went obsolete, right? So, I think my belief is that people with disabilities, actually ordinary people deserve the best technology as opposed to stuff that was, you know, from their parents or grandparents era.

LARRY GOLDBERG: Yeah. You know, I felt some pain there. I’ve been in the captioning business for many, many years, and it was very analog and manual to start. And then some smart disruptors came around and started using hybrid technology combo of, uh, machines and humans, and it was really disruptive to the captioning industry. And running a nonprofit, we had to be concerned about, can we keep up? Can we invest the necessary finances and resources to deliver? And you talk about that a lot in your book.

JIM FRUCHTERMAN: Well, I think, I think one of the things that… And I, and I think that when you get to the heart of it, the goal is to serve the people we serve. It is not to preserve the living that we have or the organization we have. And when organizational preservation starts taking precedence over the interests of the people you serve, things are going wrong. But the, you know, the way people are funded often causes these sort of divergences. But I know that the people who really innovate in a field tend to be the people who have been on the front lines, see the good and the bad of how things are being done today and try to imagine how to do it better. And I wanna help equip those people to be more successful, whether they choose to start a, a purpose-driven for-profit or a nonprofit that’s sort of more disruptive and technically capable.

LARRY GOLDBERG: Well, I think we can feel good about the fact that a lot of the young people coming into the business now are digital natives. Of course, they’re gonna use some great tools. They’re not gonna use some of the, the pedals we used to have to push to make this tech work. Now, one of the really interesting questions around being a nonprofit, uh, and trying to be a good business is that your customers are not necessarily your users. And that is a tough dichotomy. Can you talk about that a little bit?

JIM FRUCHTERMAN: Yeah, I mean, the thing about being in business is you’re, you’re c- you’re dependent on your customer paying you to be successful. And so you have an alignment. But whereas in the nonprofit sector, often we have what, you know, is called third-party payers. Sometimes happens in healthcare as well, where the people paying the bill are not the people who actually need the service. And then you get these sort of divergences. Now, you know, I think that you still should be treating your user, your community member, your beneficiary just like they were a cons- a customer. Uh, and I think that that tends to create a more positive thing as opposed to, uh, spending all your time, uh, kissing up to the funder. Um, but those are both valid strategies. I just happen to lean towards if we do the best for the person we’re serving, we can usually convince the payers that that’s best for them. And, uh, but we have to be cost-effective. We can’t just say, “Well, we could spend an infinite amount of money having, you know, a gold-plated product.” We don’t have that option. We have to exist in the real world. But I… you know, people first, uh, efficiency second is usually the way I’d like it to go, if that’s at all possible.

LARRY GOLDBERG: You talk in your book about sometimes board members can be kinda retrograde. They want you to just stick to the old mission, do the same old, same old. You’ve spent a lot of time in your life educating board members and funders to get with the program. Uh, so how’d you pull that off?

JIM FRUCHTERMAN: Well, again, you have to go back to the first principles is, you know, why are we here? What is our mission? Uh, is it organizational preservation or is it actually serving people? And I end up advising nonprofit leaders on how to convince recalcitrant board members, how to shift them, how to make the case, and, you know, I think one of the things that I’m currently advocating for, and I’m, I’m one of the advisors to, is something called board.dev. The idea that every nonprofit needs a tech person on the board just like traditionally they needed a lawyer and a finance type on the board. Because the, the need to become savvy with technology is critical. And if you have someone who comes from the tech industry, but has been educated on being respectful of the culture of the social good sector, you know, that… and, and then work on the people who are in the social good sector to begin to understand the benefits of technology, I think bringing the best of both those worlds will come up with w- whatever the next generation should be. Because standing still is not an option, and if you think about the nonprofit sector today, which is needs are going up and funding’s going down, one of the few things that I know that allows you to do, you know, more with less money is technology. And so I don’t… I wish it wasn’t the case that we needed it because it’s a crisis, but, but still, it’s a, it’s a pretty good path and I think more and more nonprofits are realizing that they have to start emulating the business sector in using technology to become more efficient- Yeah. … um, because, you know, social demand is often there. If you can serve 40% more people with the same staff, you really should try to do that.

LARRY GOLDBERG: Yeah. And so you throw efficiency into the conversation, but also you talk about effectiveness, and you give this formula for what I call, at least, the value proposition of any social good organization. And that is you take, uh, the effectiveness, the outcomes, and you divide it by resources. And, and if that curve, uh, is, is positive, you’re doing something good, which is really smart. But you need to… So the first question is, who’s gonna decide if you’re actually effective? Is it the funders? Is it the board? Is it the clients? Who decides?

JIM FRUCHTERMAN: Well, I mean, yeah, I always like the client, and, and I think that, you know, if you have a board that’s willing to go along with that journey, that’s optimal. But I, I do think, and, and I think the world we’re in now, which is a lot more open-source software and, and social media, we have a, we have the equivalent of peer review in an academic sense. So, we’re a member of something called the Digital Public Goods Alliance, and we’ve agreed to a set of principles of how we’re supposed to share our work and how we’re supposed to design around privacy and things like that. And I think that, you know, you don’t have a guarantee that, that, you know, that’s perfect, but when you have, when you have your, when you’re part of a cohort of organizations that have sort of come together with the set of shared principles, I think you tend to get better. And that’s one of the great things about the accessibility community. I think that’s what I love, is that everybody wants to help each other improve accessibility, even if you’re sort of competing in some weird way. And I, and I think that the same thing is hopefully true of this whole sector of social impact technology. And there’s enough people doing it now that it’s not like the lonely old days. And so I’m hoping that there’s a lot of cross-pollination of ideas, and the open source movement has really helped in sharing code, and… So I, I think that those are all positives.

LARRY GOLDBERG: Mm-hmm. And you talk in the book about the for-profit approach in tech, which we’re all well aware of. Think big, think global, make it scalable, dominate your market. And on the social good side, you say, if you wanna dominate your market, you will find that you’re, it’s a market of zero, really. And instead you have to make friends and share a smaller market with other people. How do you deal with that?

JIM FRUCHTERMAN: Well, I think the, the issue is, when you start your, start a tech company for social good, you wanna focus on one narrow specific problem or one narrow specific sector. So, like, I’m doing, we’re doing helplines for kids in crisis. It’s not solving all the world’s problems. It’s solving one thing that we think is important. That we’re narrowing the focus to just that. And the same thing with, you know, I ran Bookshare. When we started Bookshare, the idea was accessible books for students. So, it, it’s, it’s not, you know, it’s not everything. It’s something. And I think that allows you to, to, to narrow your focus. And then if you’re successful with that, you can then do better. But I think, you know, I think, I think there’s this, this urge to try to be all things to all people. And I see it a lot with a lot of my entrepreneur friends who come out of the nonprofit sector, is they try to boil the ocean. They’re like, “I’m not solving one problem. I’m solving five problems.” And then you come in and you look at what they’re doing and you go, “You haven’t figured out how to really do any of them really well yet.” And so I think starting a business, for profit or nonprofit, requires you to find that niche where you can outcompete everybody else and do something that’s clearly better so that you’re able to get critical mass to have a s- successful organization.

LARRY GOLDBERG: Well, and you know, as well as I do, it’s easier to get investment if you can convince people you’ve got a really scalable model. So, what do you do in the social good world? How do you get people to invest or donate to you if you’re talking about a small, narrow problem?

JIM FRUCHTERMAN: Yeah. Well, I mean, I think some of it is sort of like, like Bookshare, my idea is we’re gonna solve this for, for one country, and then we’ll move on to other countries. But let’s get critical mass in the United States or in Canada or in Britain or in India. Let’s show that we have a proven model in a large enough country that, you know, we can serve- Actually, we ended up, we served 80 countries, right? Now we’re part of the Internet Archive. They, they’re carrying that forward. But each country is, is a thing to focus on because they have different educational systems. And that’s, like, when we, when we do kid- kids’ helplines, they’re all country by country because every country has a different crisis line system, and we respect the culture of the country and don’t try to impose a Western set of values on it. We just try to help the people who are already doing good work in that, in that country, give them better technology than they’ve got. And so I think, I think most tech has a global applicability, but you, you have to sort of go step-by-step and, and earn your way in, and demonstrate that your technology is beneficial to one sector, and then you can make the case to the next sector. And I think, I think funders who are… If they’re rational, you know, they’re like, you know, the first question is, “Will this org actually do what they say they’re gonna do? Are they, are they competent? And have they actually proven that?” And as, as you begin to build a track record, it gets easier to recruit additional funders. But I agree, the first funder is always the toughest one.

LARRY GOLDBERG: Yep. Let’s pivot a little bit to your thoughts on AI in your book. We’ve heard all about the big grand AI advances that are coming from the major players. Uh, what do you feel about all this? What can we believe? What can we see? What can we actually implement?

JIM FRUCHTERMAN: Well, I’m still a tech optimist. I’m also a tech realist. I believe in keeping our eyes open for, for risks. Um, but I am, I am just, I’m one of those techies who believes that more innovation and more deployment is better than trying to stop it. And, and, and I think one of the things that we know from tech is that stuff that was, that was revolutionary five years ago is obsolete now. And, you know, and so I think, you know, AI is following that same path. The only thing we know is whatever’s cutting edge today will, will be followed by another technique, another trick, another generation. And so I think th- the question is, is, you know, having, being a user of, of technology is helpful, right? Having, being somebody who makes technology is sort of, is better. The only problem is if you become dependent on a commercial entity for one particular trick, it’s gone in a few years. And so I think, as I, you know, and so I, I think, I always like to be an advocate for, if it’s important, do we keep our options open? Can we, you know, should the community that’s using it own the underlying technology? Should it be something that’s not likely to, not likely to be yanked from underneath them? And I, and I believe in open source, open tools, open collaboration, even though there are real benefits from, for profit, uh, activity. And I think we can get some of the benefits of that in the social sector, if we’re smarter about, uh, if, if we’re smarter about what we build.

LARRY GOLDBERG: So you’re, you’re a proponent, then, of open source developments and, and a community that can kind of hold that together?

JIM FRUCHTERMAN: Yeah, I mean, I think that you have to, like, in the old days, if you made a product, you owned the product, right? And, and that’s still useful for certain kinds of companies. But I think more and more, like, one of the favorite examples in the book is, is something that came out of Stanford, you know, a, a project called PrototypeA, where PrototypeA was a, a, a set of techniques for, designed for emergency responders. And they, they spun off, uh, two nonprofit organizations out of, out of this. They, they were developing a lot of intellectual property, but they’re giving away the IP to be used for social impact. And yet, they have commercial partners who’ve licensed that IP so that they can sustain, they can sustain the underlying development. So you, so you can end up in a place where you’re, you have this sort of hybrid structure where you’ve got a nonprofit and a for profit and they’re sort of in a partnership to build things that benefit both the commercial world and the social world. And I think it’s a lot more experimental right now. But if you look at something like Creative Commons licensing, you know, which was novel 25 years ago, it’s pretty mainstream now. And I think that our attitudes about intellectual property are different today than they were 25 years ago, and I think it’s mostly better.

LARRY GOLDBERG: And it’s led to some really interesting innovations that are accessible. And, and I love that. But the big proprietary AI systems that we read about every day from the big players, do you think that these proprietary algorithms are going to leave the social good, you know, space a little bit behind? Should we worry about that?

JIM FRUCHTERMAN: Well, I mean, they, they probably are going to- I mean, if I look at what Meta did with Llama, you know, Llama is mostly open weights, even though it’s not fully open source. There are legitimate issues that have been raised about it. But the fact that we can download it, we can play with it, we can run it on our own systems. There’s a lot that’s good about that. You know, and I don’t think Meta would have done that without kind of the open source community pushing for it. And I do think, you know, Google has come out with some, some good tech that they’ve released for the community. Same with Amazon. I think that those big companies have real benefits from what’s in the open ecosystem. And so there’s this interesting interplay of, I think, a healthy interplay of competitive forces and c- collaboration. And I, and I think we have to be careful that some companies or, or, or small groups of companies don’t basically capture the market and make it impossible for anybody else to compete. But I think the world we’re in now is better on that count than, uh, certainly the software era, and I, you know, and I’m hoping that, uh, uh, that the community, uh, continues to, uh, uh, continue to move forward with techniques that are, uh, that are open. And I’m, and I’m still a fan of the nonprofit tech sector playing, playing a key role in doing the research and development that doesn’t have a, a commercial payback.

LARRY GOLDBERG: Mm-hmm. I can imagine, and I know from my own experience, that if you’re, let’s say, a company that’s using AI tools and using mostly open source tools and you integrate a proprietary player somewhere in the process and they decide to change the price or change the license agreement, you’re kind of stuck. So, those are some of the dangers-

JIM FRUCHTERMAN: Yeah.

LARRY GOLDBERG: … that you warned about in your book. That being hostage to a, uh, a big player is not good if you’re trying to deliver good service.

JIM FRUCHTERMAN: Well, and I think, I think, I think as, I think as a field, we’ve learned from the Microsoft and, and Google experiences where we’ve been sort of herded into particular, uh, uh, technologies. And I, I think that, you know, and, and there’s some value in saying, “Look, maybe it’s not cutting edge, but it’s, but it’s usable, and it’s gonna be around, and we own it. Like our, our community owns it. It’s ours.” And, and sometimes that is an advantage over, uh, uh, over the, the, the, the latest and greatest. Because if you can’t, if you can’t sustain it, if it’s something that’s, uh, that’s, that’s at the whim of one company’s product strategy, uh, it may not, it may not be as good. And I think people have learned that lesson in a variety of fields and they’re, they’re trying to, uh, not leave themselves so vulnerable.

LARRY GOLDBERG: Yeah. Now, I, I imagine that at some point in our conversation, you’re gonna get asked about what, what do we expect of government in all of this? Mm-hmm. So let’s go there now.

JIM FRUCHTERMAN: (laughs)

LARRY GOLDBERG: Are we expecting that government will adopt and use some AI tools? Are we expecting they’re going to regulate? Are we expecting they’re going to get in the way? What, what are you thinking about government and business?

JIM FRUCHTERMAN: Well, I think there’s a lot of pivot going on. I’m certainly doing it and not look to Washington, but look to the states. These days, look to the states. I live in both Massachusetts and New York. I’ve got two lovely bubbles I live in, and they are making movements, and we can feel good about what’s going on there, and that can be replicated as, as Colorado and California and many other areas in, in the US.

LARRY GOLDBERG: Yeah.

JIM FRUCHTERMAN: And, and I think the, the thing is that you also have conservative states that are often doing some cool innovation, not, not necessarily bec- inspired by, I don’t know, more, more liberal objectives. This is like, “Hey, this is the job we gotta do, and if we can do it better, we will, you know, and for less money, let’s do that.” I mean, there’s a… Well, one of the startups I talk about in, in the book is a group called Recidiviz that works with Republican and Democratically-led states and says, “You know, we can save you money and preserve public safety by getting a lot of people out of prison. And y- you already have the data and you already have the laws. We don’t, don’t need to change anything. We just, you just, you just ha- you just have to take advantage of the fact that these people could be at a lower level of incarceration that saves the taxpayer money.” That’s a popular message to a lot of people.

I mean, I mean, I always had really good luck selling, um, Bookshare on Capitol Hill to Republicans and Democrats. Why? ‘Cause we had 1,000 kids in every s- every congressional district, and everyone… Well, a very small minority didn’t believe the government should do anything, right? But, but an awful lot of people are like, “Hey, kids with disabilities deserve an equal shot, and you’re doing it 10 times more cost effectively than books on tape sent through the mail. So yeah. Let’s go for that.” So, I don’t know. I’m, you know… But yes. Often, you can, you can make a lot of progress at the state level when things on, in federal land aren’t really moving forward that fast.

LARRY GOLDBERG: And you just talked about some of the data you use to convince people. Can we rely on data still? I’m a little worried about, “That’s your data.” “No, that’s not my data. I don’t interpret it that way.” Is there a ground truth?

JIM FRUCHTERMAN: You know, it’s, it’s this weird sort of thing. It’s like, if, if you’re in tech and you’re using data to figure out what it costs you to acquire a customer or what your lifetime, you know, rewards for having a customer, no one debates whether that data’s valid ’cause their livelihoods depend on it. So we have this culture where, you know, data in business is holy (laughs) and where data in science is increasingly questioned, right? So, I don’t know. I’m, I’m a data guy. I mean, d- data’s got its pros and cons, but, uh, I mean, you know, in God we trust, but everyone else, bring data.

And the data you’re talking about that is in business is really what the for-profit people will never turn their back on. Look, my bottom line, my quarterly results. I know what I got.

LARRY GOLDBERG: Yeah. What- And- … worries me sometimes is saying, “Well, if you make your website more accessible, it’ll improve your bottom line.” It’s like, I haven’t seen that data. You know, some of these things are hard to do, and some of them you do for civil rights reasons. But we always, we always try to make both cases, and sometimes that case works. Sometimes the data is elusive. And, but I think in general, more data that is more authentic is scoring with the majority of policymakers.

So, uh, you know, I mean, right now we operate 911 for kids in 20 countries. And s- ’cause often in the country, there, there actually is a, a 911 number, three or four digit number that is for kids in crisis. And those helplines, because we’ve given them, we built the technology that they run their, their helpline on, that they now are the authorities on what things kids are running into in the country, and they’re independent of business. And policymakers are a little leery about self-reported data from the tech industry, and they have a reason to because, you know, if you have an interest in what data you report, you may not report all the data you actually have. And so having an independent check on, well, gee, cyberbullying seems to be a much bigger issue when kids call in in crisis than industry is telling us.

So I mean, I, I do think that more data, carefully framed, not overstated as representing more than what it, than what it says, is better than flying without data.

LARRY GOLDBERG: Well, and that authenticity and trust, and trust in particular is something that you raise is essential for anyone trying to start a business, nonprofit or for profit.

JIM FRUCHTERMAN: Yeah, but it’s especially true of the nonprofit sector where the currency is more trust than money. Um, and, you know, I think the… I mean, the great thing about the accessibility industry is it, it does run on trust, and when people lose the trust of their customers, they get themselves in deep trouble. We have a tech industry that is dominated by people who are not trustworthy, and it doesn’t seem to be hurting their bottom line. And that is making them feel pretty darn, um, invincible about continuing to exploit our data, and I’m hoping that the, the good guys and gals are, you know, going to get, make more progress by having more trustworthy products. And I think, you know, just to make an example, Apple gets more credit for being, you know, protective of your privacy in certain ways that other people aren’t. And I, I love it when people are doing that. And I hope the rest will see the chickens come home to roost. It feels like that’s going to happen.

LARRY GOLDBERG: So you close your book with a clarion call to go do tech for good. Um, and, uh, and we of course have been doing that our whole lives. Beyond AI, what’s your advice as closing words for people who are thinking, “Okay, what’s next? What can I do to have a good social impact?”

JIM FRUCHTERMAN: There are so many things that are out there that aren’t gonna make you a billionaire, right? And so, and I joke-

LARRY GOLDBERG: Hey, a million would be fine.

JIM FRUCHTERMAN: You know, but you, a lot of people want to move from big money to meaning, to big meaning, and make a living. And so, so, uh, what I tell people is, you know, acquire the skills of whatever is your, of your dream, right? Whether it’s a technologist or product designer or marketer, whatever it is. Acquire those skills, and then when you have a dream of something that could actually make society better, make humanity better, make the planet better, uh, and you figure out it’s not gonna make me a boatload of money, you don’t give up on it. You actually sit back and go, “No, wait a minute. I need to get a copy of Jim Fruchterman’s book to tell me how to actually do that.”

But, but there are hundreds of examples of people who have managed to make this pivot, make a good living, and make a huge difference in the world. And I think that is the accessibility industry. You know, the choice between for profit and nonprofit is not a moral choice. The choice is what’s the most effective way to go about this for your, for your objectives? And often that’s gonna be a for profit, but do it. Do it no matter what. Don’t give up on it just because there isn’t a billion dollars in it.

LARRY GOLDBERG: I love hearing that. Now, did I see you got the book with you? You want to hold it up?

JIM FRUCHTERMAN: Yeah, I do. Here we go. Let’s do it. It’s, you know, Technology for Good. All right. Uh, first we need the fuzzing algorithm and, uh, there we go. We got the angle on it. But, um, and it’s, uh, in an accessible ebook, certif- already global certified accessible. And it goes open source in a year.

LARRY GOLDBERG: Fantastic. Fantastic. Well, thank you for your time, Jim. I always love talking with you, and we could go on for another 10 hours. But for today, we’ll say goodbye and thank you to you and to Sight Tech Global.

JIM FRUCHTERMAN: Sight Tech Global rocks.

[MUSIC PLAYING]