The miracle-working CEO of portfolio company Finetune Learning, Steve Shapiro, is back to tell us about the latest developments in applying artificial intelligence in educational assessments. We also got into remote work and explored how to support younger workers in creating professional networks while working remotely.
Highlights:
Sal Daher Expresses Gratitude for His Sponsors: Purdue University Entrepreneurship and Peter Fasse, Patent Attorney at Fish & Richardson
Sal Daher Introduces Steve Shapiro the Miracle-Working CEO of Finetune Learning
Finetune Learning Uses Technology to Support the Creation of Great Educational Assessments
Version 1.0 of Finetune Made It Easier to Grade Essays Consistently – A 2X Improvement
Finetune Version 2.0 Is More B2B – It Is Assisting in the Effective Creation Content for Educational Assessment
Making Assessment Writers 10X More Effective
“The other product that we have that scans the text, then classifies and tags it, does it in two orders of magnitudes more than a team of humans. Let's throw away the 2X to 10X and start saying 100x, 200x.”
“...we never said this is something that replaces a human. We believe that strongly. We always say this is something that's going to augment humans who are really good at this.”
Pre-Trained Systems Will Revolutionize Adaptive Learning
“In most cases, we're not a full solution. We are either a widget or an intel inside a larger solution with a large partner that's a multi-billion-dollar company.”
"I went into this because I actually hated school and you're throwing me back into a classroom....”
Old-Time Adaptive Learning
“That missing 10%, they have to be trained and they're not there because the teaching methods don't fit their learning style.”
“...there's a ton of Gen Z and Millennial founders out there that are starting to create companies in education technology...”
Unexpected Source of Talent for Cybersecurity
Finetune Driven to Remote Work Before the Pandemic Due to the Difficulty of Competing for Talent with Amazon and Google in Kendall Square
Lessons on Managing a Remote Workforce – Learned from GitLab
The Right Approach to Company Culture
How a Founder Learned to Get Along with Her Co-Founder
How Younger Workers Are Disadvantaged by Remote Work
How to Help Younger Workers Build Social Networks? A Question for Jacqueline Olds, M.D.
How Do We Reskill Angel Investors to Help Build Life Science Startups?
A Happy Outcome from Zoom School
Steve Shapiro’s Parting Thoughts
ANGEL INVEST BOSTON IS SPONSORED BY:
Transcript of “AI in Education”
Guest: Steve Shapiro
Sal Daher Expresses Gratitude for His Sponsors: Purdue University Entrepreneurship and Peter Fasse, Patent Attorney at Fish & Richardson
Sal Daher: I'm really proud to say that the Angel Invest Boston podcast is sponsored by Purdue University entrepreneurship and Peter Fasse, patent attorney at Fish & Richardson. Purdue is exceptional in its support of its faculty, faculty of its top five engineering school in helping them get the technology from the lab out to the market, out the street, out to the clinic.
Peter Fasse is also a great support to entrepreneurs. He is a patent attorney specializing in microfluidics and has been tremendously helpful in some of the startups in which I'm involved, including a startup that came out of Purdue, Savran Technologies. I'm proud to have these two sponsors for my podcast.
Sal Daher Introduces Steve Shapiro the Miracle-Working CEO of Finetune Learning
Sal Daher: Welcome to Angel Invest Boston conversations with Boston's most interesting founders and angels. I am privileged today to have with me, again, an alumnus of the podcast, Steve Shapiro.
Steve Shapiro: [chuckles] Thank you, Sal. It is such a great honor. It's a huge honor the first time, but I can't even describe the second time, yet.
[laughter]
Sal Daher: No, Steve, let me tell you something, Steve's podcast is a hugely popular podcast because it's titled, Miracle-Working CEO, which he is. He's a humble guy but he really does reserve that title. He's done amazing things with Finetune Learning. Steve, tell us for our audience who has not yet listened to your backstory, and I highly recommend that they go and listen to it in Miracle-Working CEO, the podcast, the original one, tell people what Finetune Learning does and why it's really important? What problem is it solving?
Finetune Learning Uses Technology to Support the Creation of Great Educational Assessments
Steve Shapiro: Thank you. Yes. Thanks, Sal. Finetune Learning's an education technology company that from its founding days was always steeped in the idea of learning, really, is always in need of great assessment, and that assessment is an often-maligned term in the world of the learning ecosystem, but an extremely important one, and we always felt there was a better way. We pretty much pioneered a new assessment delivery system that looks at what is it like to be able to look at student work that is subjective and give them consistent and reliable feedback pretty much in real-time when the topic in the question is subjective. This is not a multiple-choice question. This is not true or false.
Sal Daher: You're breaking students, not only students that are only kids, people should be learning life long, we're going a to get into this, later on, you're breaking people who are learning out of the old constraints of multiple choice.
Steve Shapiro: Exactly.
Sal Daher: Now you can have people giving answers that aren't structured. Finetune Learning can make sense of that.
Steve Shapiro: That's right. We've evolved as a company. Part of what we'll talk about today also is just some great lessons learned as to how we've evolved, as the industry has changed, as technology has changed for all of us. It's a great marker that I actually joined in late 2014. This is a seven-year journey at this point. When I think about when I joined to now how much actually just the world has changed, I'm actually putting aside, obviously, the pandemic, but thinking about just, in general, how technology has evolved, how it's been adopted or not in certain parts of our life including education, and how we as a company responded to that, we've really evolved as a company and our product has evolved in response to all those changes.
Sal Daher: Now, Steve let's get back and explain to listeners what the product offerings of Finetune Learning is?
Version 1.0 of Finetune Made It Easier to Grade Essays Consistently – A 2X Improvement
Steve Shapiro: Sure. The product that we started with seven years ago was a digital platform for assessment that did exactly what I said earlier. It allowed teachers to, one, get trained, "How do I look at student work that is subjective so that I and my colleague have some consistency with the way we treat it and score it and give feedback on it?" Think about every person when they're in high school and their friend has the easy English teacher and they have the difficult one, [chuckles] this is trying to solve that problem. This goes back 10 years ago to our founder, Ogden Morse, and I give him a lot of credit. He was an English teacher so he was the persona of the problem we're trying to solve.
The early platform was a human-enabled digital workflow, and that was great, the human-enabled digital experience for teachers to get calibrated in a way of looking at standards that are embedded in a rubric. The rubric is, this is how we evaluate student work. You, as a teacher, first, have to prove out almost that you're consistent with and reliable with the way you use that rubric as compared to a panel of your peers. That's how it started. Over time, and in response to the way the world has changed, we saw a lot of other people going a few steps further than that. We saw people saying, "Well, where else can we put automation in it?"
About two years ago, what we stumbled on was not where can we continually automate moving forward from there, because moving forward from that would be, "Oh, can you use some kind of AI to grade it, some kind of AI to give the student feedback?" There's a hundred companies doing that now. We saw, actually, another problem and a problem we thought that we could actually incubate the technology to do. That's what our new product is about. That is basically, again, using the latest innovations in AI, transformer networks.
Many of our listeners have probably heard of GPT-3 and the different versions of that and leveraging that in a very interesting way to create content, specifically, for us, starting with creating assessment questions, and secondly, how to classify and tag content. Classifying and tagging content using AI unlocks an amazing bevy of potential products for everybody in the learning ecosystem, including education publishers and anyone who has content, including an individual instructor.
Sal Daher: Let's get concrete. Let's say you are a harried English teacher who is using your platform. How does that improve your life?
Steve Shapiro: Yes. Using the 1.0 version of the platform, it improves your life because it allows you to take a data-driven way of looking at student work that's subjective. Prior to using us, you could never look at almost data-driven, but ours is yielding these types of analytics that you never could have thought of before because, again, it was subjective. You're grading a student essay, you marked it up and you gave it a grade.
This creates more structure to it so you could really say, "Hey, my entire class really struggled on that part of the rubric. Either it's not getting through to them, I didn't teach it properly, or I taught it properly, but a lot of people are still struggling with it, so let's not move on. Let's circle back and make sure they understood that first." We hadn't had that data-driven approach to what is inherently subjective before. That's the 1.0 version, the 2.0 version—
Sal Daher: Hold on a second. The 1.0 version, what it's doing, it is empowering the teacher with tools that businesses have had for almost a full generation now. People who are on the production line and quality assessment experts and so forth, they have all these tools to keep track of what's happening with the process. Version 1.0 was providing that for the teacher in the classroom without adding a lot of burden.
Steve Shapiro: That's pretty well put.
Sal Daher: Okay, so next version 2.0, what's going on there?
Finetune Version 2.0 Is More B2B – It Is Assisting in the Effective Creation Content for Educational Assessment
Steve Shapiro: Yes. Version 2.0 right now, for us, is more of a B2B product. We'd be working through large education publishers and assessment companies and test prep companies or anyone who's doing any form of assessment, but in the end, their users will be able to use this product to say, "Hey, I want to actually create some new assessment questions. I don't want to rely on what's out there. I don't want to rely on this textbook. I want to create brand new ones and so I'm going to have a set of dropdown options of- I want it to be related to this learning objective for my students. I want it to be this level of difficulty."
Making Assessment Writers 10X More Effective
All of these types of things to filter and then, boom, I'm going to press a button and a pretty darn good one is going to come out and I'm going to QA it because I'm the subject matter expert. If I don't love it, I can tweak it, press the refresh button again and, boom, it's going to do its thing. Over time, I'm going to be able to write 10 times as many questions as I could in whatever time it took me to write one question and it's going to unblock me from writer's block because it's going to be spitting out creative questions. That's only half our product by the way.
Sal Daher: Okay. No, no, but I like this number already because in our first interview, you were 2X-ing the ability of teachers to grade papers. Now you are 10X-ing the ability of people who are writing questions for assessments.
Steve Shapiro: That's right.
Sal Daher: This is like taking off.
“The other product that we have that scans the text, then classifies and tags it, does it in two orders of magnitudes more than a team of humans. Let's throw away the 2X to 10X and start saying 100x, 200x.”
Steve Shapiro: You're setting me up for a humblebrag, but I'm just going to say it. [chuckles] The other product that we have that scans the text, then classifies and tags it, does it in two orders of magnitudes more than a team of humans. Let's throw away the 2X to 10X and start saying 100x, 200x. We'll get into this later. That, all of a sudden, gives everyone we are working with the ability to do things they never dreamed of doing before, so we think it's of incredibly high value.
Sal Daher: I was happy with 2x. I was ecstatic with 10x, and I am stunned by 100X. 10², this is awesome. Give me some more detail on this.
“...we never said this is something that replaces a human. We believe that strongly. We always say this is something that's going to augment humans who are really good at this.”
Steve Shapiro: Sure. Like any product, we're on a product journey with it. We are about two years in. Along the way, we never said this is something that replaces a human. We believe that strongly. We always say this is something that's going to augment humans who are really good at this. Whether you're an instructor or a content person working for a publisher, it's going to augment you. It's going to allow you to write a lot more questions faster and better and quicker and more creatively.
If you are a person who also is in charge of, "Hey, we're swimming in a sea content. How do we make sense of it? How do we see how this content relates to different classification [of] things out there, taxonomies out there? How do we do that?" Again, it would be a team of people doing that, but this would allow the team to do it a lot faster and not only do it faster, but because they can do it so much faster, they could think about what are other taxonomies this aligns with?
Just to give you an example, most higher education textbooks-- What is most students' path after graduating from undergrad? It's to start their career if they're not going to grad school. How often you, in the middle of a course or in the middle of a textbook, and it's been tagged to 100 different industry taxonomies for the jobs out there, career paths out there, things like that. That's never actually been done before. It could be done. That's a huge human labor, endeavor.
Sal Daher: What you're saying is that, let's say, you are a college student and the content in your textbook is going to be written to address topics that might be in a dozen or two dozen different applications that you might encounter after school.
Steve Shapiro: Yes, I think that's one part of it. The other part of it is, as we live in a world of digital learning and your textbooks and e-textbook is-- Maybe that chapter of Econ 101 looks the same to you as it did for you and I, however, there's a bunch of hyperlinks in it. [chuckles] Those hyperlinks are helping you understand, "Hey, you're not just learning this about monetary policy."
If this interests you and makes you passionate about, "Is there some job out there or some industry that I want to be involved in monetary policy?" well, this thing could be the pathway to, "Oh, by the way, for people who master this set of skills, here is a career path around that. Here are the different industries where this is considered very relevant and in demand."
This an almost unlimited potential for the way that people who do deliver learning, whether it's online courses, whether it's textbooks, whether it's boot camps out there, anyone in the learning ecosystem would want to be able to say to a student who's in the journey of that particular course where can you go with this if you're really stoked about this chapter.
Sal Daher: This is interesting. I suspect that if you are doing 100X what human beings used to do, you're going to be breaking the problem that we have with-- Let's say that you are an aspiring data scientist and you go to a data science program at a university, you're actually learning stuff that might be five years out of date. What's going on in the industry is ahead of that but somehow it doesn't filter back.
I wonder if there isn't a way that by enabling people who are writing content for courses to make sense of much more material. They might just actually becoming aware of stuff that's going on, it might give them the warning, "Hey, your curriculum, there's other stuff that should be here. I know that you're talking about this, but if you're talking about this, have you considered this, and this, and this?" The way that diagnosticians, physician, they diagnose with some of these systems, the way they're going right now is, "Yes, I know your theory is that the patient has X, but have you thought about Y and Z?
Pre-Trained Systems Will Revolutionize Adaptive Learning
Steve Shapiro: Exactly. Just to be clear, I've been involved in the world of adaptive learning since it really started, which is about 2005, but not really until about 2012 to be honest. This is relatively new, but it's gigantic now and most education technology strives to be adaptive. by adaptive, we mean, is there something about this experience for the student that is personalized to that student based on something. Usually, that something is how they answer formative, short assessment quizzes as a way of indicating to the machine their level of understanding or not understanding, and there's some journey and path based on that.
Adaptive learning has had fits and starts, and it's evolving, and it's not there yet. Part of the reason it's not there yet is it requires a level of tagging, a deep level of tagging of the content for the recommendation engine to truly say, "Okay, based on these last three or four assessments and the way I see you've performed, I think you need to go down this path." There's a lot of decent adaptive learning out there. I think it's easier to architect for certain subjects versus others, obviously. It's a little easier to do it with technical stuff like math and science but still, to do it really, really well it's the tagging. It's the tagging in the classification.
Sal Daher: How is that tagging and classification going to be accomplished?
Steve Shapiro: Oh, you mean by us? [chuckles]
Sal Daher: How are we going to train the machines? The stuff that machines have to learn is really stupid, dull stuff that to humans it's kind of obvious. How is that going to happen?
Steve Shapiro: Sure, without giving away the secret sauce on this show, which I'd love to, Sal, patent-pending and everything, but I think we are using transformer network technology GPT-3, the P and the T is for pre-trained. We know that these giant AI engines are pre-trained on huge corpus of content on the internet, so in some respects what we are doing is we're working and partnering with those giant engines to make sense of what the output is.
One of our advantages is that requires actually less pre-training. I don't want to minimize the amount of effort that goes into what we do but it is less pre-training. That's a huge advantage, is that we're working with models that are pretty amazingly trained.
Sal Daher: What is a transformer network? You've used the word twice. It's got over my head twice. Can you explain it to our listeners.
Steve Shapiro: Now you're really putting the non-technical CEO on the [chuckles] hot seat. About two years ago, right around the time the pandemic started actually, you started reading about transformer networks and GPT-3.
Sal Daher: This is where I should have a hyperlink to Gil Syswerda, my buddy, who's a pioneer in AI.
Steve Shapiro: The pre-trained networks are gigantic endeavors. The most famous one is known as open AI, and its investors who are Microsoft and Elon Musk, and literally a $1 billion is put into training the first model. It, basically, ate the internet in some respects.
Sal Daher: Okay, so it's ingesting huge amounts of unstructured data to try to make sense of that?
Steve Shapiro: To make sense of it, exactly. There was a GPT-2 before GPT-3, and this is the next version which is more powerful, and like all things in technology, leveraging the fact that- computing power, Moore's law, whatever. It's always going to get better and bigger and stronger, and in this case, actually more real and more precise. The first media about it that came out from MIT journal and different tech magazines was, "Hey, beware." This stuff can write an article, and you wouldn't know that it was AI versus a human, right?
Sal Daher: [laughs] Yes.
Steve Shapiro: People are saying, "Hey, this could be disruptive, fun, social media, and all sorts of things." All of those things are true. I'm not here to belittle any of those things but the point was this stuff is becoming more indistinguishable from a human. It's drawing inferences that most AI has not been able to do in the past or the type of inference that you could draw would have required so much pre-training that it almost became too difficult to do.
Sal Daher: Gil Syswerda thinks that artificial general intelligence is not too far off. When that happens, it's going to take off because the AGI will be working on creating AGI too, which is more sophisticated, understands more. We're just at the brink of something really extraordinary. You double the ability of a teacher to a correct paper, then you 10X a question writer's efficiency, and now you're talking about 100X, the understanding that people have of various subjects, and putting them in context. We talked about topics we wanted to touch on. Where does this fall? Is this like the changing world of learning, how learning is changing?
Steve Shapiro: I think it falls into part of that category. I wanted to talk about the changing world of learning. A lot of it has to do with a few of the things I already mentioned around adaptive learning and the ability to get more out of content and have more context. You know, we've always been dedicated. In the end, we want to be the people that help instructors and support instructors' student outcomes improve. That's a pretty simple mission. It's a good mission. It was an easy mission from the start with our founder being a teacher.
“In most cases, we're not a full solution. We are either a widget or an intel inside a larger solution with a large partner that's a multi-billion-dollar company.”
The technologies evolved... we're saying, there are certain places that we have expertise and we want to play. In most cases, we're not a full solution. We are either a widget or an intel inside a larger solution with a large partner that's a multi-billion-dollar company. We think that the uniqueness of what we're doing is unlocking a lot of what I mentioned to you before we get on this call, just around the way the world of learning is evolving.
Sal Daher: Do you want to put some bullet points just to structure that thought a little bit?
Steve Shapiro: Sure.
Sal Daher: How is the world of learning changing and how does it affect people in general?
Steve Shapiro: Part of a network in our industry that always attends our big industry event every year in our education technology is the ASU GSV conference. GSV, Global Silicon Valley Bank and Arizona State University have created this community of around 10,000 of the major players in education technology that convene at San Diego every year, and also just part of a community. They're very smart people and one of the adages is they started seeing about our basically learning environment or early learning ecosystem is we're servicing pre-K to gray.
It was this recognition, I don't know, the last 5 to 10 years of, A, "I know a lot of us grew up thinking when I graduate from high school or I graduate from college, I'm done learning. I have a job," that kind of mentality. I'm maybe dating myself, but born in 1962, but the world has changed so much. That's not the attitude you want to have, and that the attitude is, "How can I continually engage and continually grow, and have that growth mindset, as they say, around learning because it's going to really benefit me. It's going to benefit me from a lot of different perspectives, not just career.
The obvious one is it's going to benefit my career as I continually evolve and learn more and have a subject matter expertise and things but being intellectually curious your whole life has been proven to be a total health boost, your personal health and your mental health. It's proven. That's a big component, is this idea of lifelong learning, but lifelong learning has to be, for those of us who are really fortunate and could take any online course in the world or join any different place that's doing something interesting, it has to be more accessible.
That's a really interesting way we think about how is this more accessible? How is this more accessible to less privileged members of our society? To our credit in our industry, a lot of people are really thinking about how to solve those problems. We definitely are at our company as well.
Sal Daher: How do the machines help here? We talked about one way is putting assessment into the classroom so that the teacher can understand based on the curriculum that she or he has, how the students are learning or if it requires more attention, and so forth. We're seeing that in the classroom, but how do you expand it beyond that? What closes the loop for someone who is, say, picking up a course to keep on top of the particular profession they're in, or someone who's a machinist, or somebody like that.
Steve Shapiro: Not to dodge your question, because I'll answer it, but one of the things that always really struck me was for a lot of the last 10 years when you'd see an article about how technology is displacing some jobs and-- We go to these conferences and say, "What are we going to do to people who have more, what used to be called, blue-collar jobs, if those jobs are disappearing? How do we reskill/upskill?"
"I went into this because I actually hated school and you're throwing me back into a classroom....”
There was this article about some, I think, it was a Ford or GM plant in Wisconsin and they were like, "Okay, so we're closing the plant, but we're going to all reskill you." I don't even remember what they were reskilling them in, but what everyone who was in the class was saying was, "I went into this because I actually hated school and you're throwing me back into a classroom. I'm 45 years old and I've been doing this for the last 20 years. It's the absolute, it's giving me PTSD."
I think the answer to your question is what are the better ways that, one, we can make better learning accessible and what are the ways to do it in a way that's not giving people PTSD. It's not all solved, by the way, by AI or by computer. I'd say just the opposite but the AI is the helper. The AI is the helper. That's where we stand. We think there's so many situations where-- What's going to be beautiful about this is there's an instructor or facilitator, who's able to help people on a learning journey or people are doing group work and that's part of the learning journey. Somewhere embedded in that is something that's automated that's helping them and adaptive learning is part of that.
Old-Time Adaptive Learning
Sal Daher: It's funny, adaptive learning, when I was a kid, it took the form of these boxes that had these reading sections that you progressed on them, you took a little test, multiple-choice tests. You went along and so forth and some of them were purple, and orange, and yellow. Then I actually met the guy who ran the company. His name was Frank Ferguson, and that's Curriculum Associates. I was like--
Steve Shapiro: I knew him well.
Sal Daher: Knew him well, the late Frank Ferguson, and what they were able to do-
Steve Shapiro: Amazing man.
Sal Daher: -with just card files. My grandson is in Montessori school and he comes home and he says, "Oh, I just finished the Waseca drawer. I'm on the orange Waseca drawer," and so forth.
Steve Shapiro: [laughs]
“That missing 10%, they have to be trained and they're not there because the teaching methods don't fit their learning style.”
Sal Daher: These little devices, I cannot help but-- One of the problems we have is that right now college graduates are something like 60% of college graduates are women and then only 40% are men. What's happening is that a lot of those men who are not showing up at college, they need to be trained. We have a training system that is geared towards high school instruction, college instruction. Men and women in the population are represented about 50-50. You would expect the college population to be about 50-50, right? That missing 10%, they have to be trained and they're not there because the teaching methods don't fit their learning style. You need much more hands-on. I don't want to be a Metaverse fanboy here. I think that's there's a lot of baloney there but do think that there is potential for use of enhanced situation visually so that you can see things or visualize things for people who might be visual learners, instead of people who learn from text.
Steve Shapiro: That's right. Sal, you know that I'm a venture partner at LearnLaunch, the edtech accelerator in Boston.
Sal Daher: Oh, you are?
Steve Shapiro: My friend, Jean Hammond, who's also your friend.
Sal Daher: I know Jean Hammond, alumna of this podcast, honored alumna of this podcast.
“...there's a ton of Gen Z and Millennial founders out there that are starting to create companies in education technology...”
Steve Shapiro: Yes. Big shout out to my buddy, Jean. One of the things about being involved there is that, obviously, people see that you're a venture partner there on LinkedIn, and I get about 10 outreaches a week of people who are like, "Hey, I have an edtech idea. Can I brainstorm with you?" Selectively, I can't talk to everybody but those that I do, I learn so much from them. I'm here to tell you that what you just said about different ways of learning, there's a ton of Gen Z and Millennial founders out there that are starting to create companies in education technology, with the main premise is we all came up on something that is resembling TikTok. TikTok is how we engage. Learning in a TikTok style of short video, entertaining video, there is something going on in the background that's serving you up the right, maybe not like--
Sal Daher: Algorithm-driven.
Steve Shapiro: Yes, algorithm-based.
Sal Daher: Driven by the brilliant algorithm.
Steve Shapiro: Exactly, yes. Sometimes it's off for me. I haven't quite groomed it yet but in general, there's a lot of ideas out there that I think are very targeted to this current generation of people who are in their teens, 20s, and 30s that understand like, "This is the way we learn," which is very different from the way you and I were taught. I say bravo and viva. Yes, right?
Sal Daher: I can imagine TikTok-like instruction videos for plumbing for someone who's going to be a plumber.
Steve Shapiro: Absolutely.
Sal Daher: That is so exciting. That is really exciting. Human capital is the scarcest thing there is-
Steve Shapiro: Don't get me going. [chuckles]
Sal Daher: -for it to be wasted because we're not developing it properly. It's such a shame. [crosstalk]
Unexpected Source of Talent for Cybersecurity
Steve Shapiro: It's interesting that you mentioned plumbers because about three years ago, I was helping a friend who was trying to do a cybersecurity training venture. One of the things that we learned-- Everybody knows that this has been beamed up by the Department of Labor for about the past 10 years, that there's anywhere between 1 million- I heard recently 3 million job openings in cybersecurity in the United States.
Sal Daher: Oh, in cybersecurity, just the cybersecurity, yes.
Steve Shapiro: Just cybersecurity.
Sal Daher: Overall, it's 7 million.
Steve Shapiro: Yes, so see, I'm even-- I heard of 3 million 3 months ago.
Sal Daher: No, no, no. I can believe that it's 3 million in cybersecurity, just in cybersecurity.
Steve Shapiro: Yes, 3 million in cybersecurity and growing, and they just can't find enough people. One of the things that we learned was some large nonprofits started to do research based on the premise of, "Okay, so we're not going to get enough people out of just, say, comp sci programs. There's not 3 million people graduating from computer science, first of all, that are going to just go into cybersecurity because there's too many other options. Where are we going to get these people?" They listed the different places. Of all the different majors and vocations that they put together, plumbing was number two on the list. You say to yourself, "Hey, wait a second."
Sal Daher: Plugging the leak.
Steve Shapiro: Problem-solving, right?
Sal Daher: Yes.
Steve Shapiro: Plumbing is problem-solving, multistep problem-solving.
Sal Daher: Exactly. Yes.
Steve Shapiro: That's an example of-- How would you get people who-- You're not going to get someone who's currently being a plumber because they're way too much in demand, like, "I'm not going to take that pay cut. I'm not going to take that pay cut." You could get the person that was interested in maybe going into plumbing that maybe can't break-in, to say, "You know, maybe a cybersecurity training program and try taking a crack at that kind of career might be really good."
Sal Daher: That is fascinating. This is fascinating. This is so weird, the connection like that, that they are plugging the security leaks in computer systems, the people who are plugging the leaks in plumbing or who might have been plugging the leaks in plumbing. This is so fascinating. Steve, let's amble on to remote work. What are your thoughts on remote work?
Finetune Driven to Remote Work Before the Pandemic Due to the Difficulty of Competing for Talent with Amazon and Google in Kendall Square
Steve Shapiro: Yes. It was surreal when the pandemic started for a lot of reasons for me, but one of the reasons was that we had been a remote company for three years prior to the pandemic. We had, in 2015 or so tried to really establish a base in Boston, where I'm based and a few of my key people are. Coming to the realization that we were hiring software engineers, a few data scientists, we kept getting outbid by the big brand name tech companies. I think Google, Microsoft, Amazon were all opening offices in Kendall, and so on and so forth. It became tough for a small company to compete.
Our CTO wisely said, "We should just become a remote company and start looking for talent all over the world." As I look back, it was the greatest decision. I was old school at the time. I'm like, "Are you crazy? I want to see them every day." All the things that an old outdated person like myself might think back in 2015, but flash forward, and obviously when the pandemic broke, it didn't affect us because we had already been doing that.
Lessons on Managing a Remote Workforce – Learned from GitLab
Here's what I've learned along the way. First and foremost, yes, it's true it is more challenging. Relationship building face-to-face is far superior no matter what, and you lose that, obviously. The massive advantage is you can scour the planet now and we've got amazing people in places as far-flung as Kazakhstan. We have people in 15 different countries and we celebrate the diversity of our team. That's actually a huge drawing card to our team, that we have people in 10 different states in the United States and 15 different countries.
We definitely had some hard lessons along the way, but we were lucky. One of the reasons we were lucky is we got a chance early to connect with the guy who's the founder of GitLab. GitLab is not GitHub, just to be clear, but there's something a little similar. At the time that we met him, they boasted 1,000 people in 200 countries. They'd been doing remote work, obviously. They had actually open-sourced their HR manual. The guy was gracious enough to spend a few hours with us, helping us sift through it and understand it, and really ground us in some of the best practices of what you have to do. That really helped us.
Also, here we are five years later when we first started doing it. I would be lying to you if I didn't say we made a ton of mistakes early and we learned from them. We take for granted that when work gets really hot and heavy, and that has happened almost all the time for us, you want to sacrifice that time that was set aside for, "Hey, let's all just meet in Zoom meeting and yuck it up and joke around and not talk about work," or have a happy hour where you could have the beverage of your choice in your hand at the end of the day and decompress.
We found ourselves saying, "Well, that's the best practice that we learned because we're not going to be in the same office, we're not going to be able to go out to the local restaurant or bar after work together, but what are these things that we could do virtually that are a substitution, maybe not as great, but still good to become friends and get to know each other better?" That was one thing that we went through this period where we were so slammed and we kept not doing it for a while and it showed up. How did it show up? We lost some people. We had had a really good track record of really low turnover and everything like that.
I'm not saying that's the one reason that we lost people, but it's a contributing factor. That's 1 example of 100 things that you have to do. It's not just some discreet, "Hey, we're going to have a meeting where we're not going to talk work and we're going to have a beverage," it's more, how do you handle every meeting? How often do you meet? How do you handle Slack and the way you communicate on Slack? How do you handle the way everybody communicates? We have to work through that. We had to do it in a way where we really listened to our team. We evolved into having a system that, I think, is pretty darn good and it respects what the team wants.
We're a very team-oriented company. As a result of that, I could look and say, we're about 62 people right now, we have the most amazing team. I will wax poetically and brag about each and every one of them. That would dominate the rest of this call, but I'll just say they're amazing, but they're not amazing by chance, they're amazing by how we evolved to being the company we are, competing on things that sometimes aren't always the compensation.
Obviously, we try to be competitive with compensation, but in this day and age as a tech company, it's really darn hard to compete with Jeff Bezos and Mark Zuckerberg, the way they pay software engineers and data scientists and stuff like that. You have to compete on other things. One advantage we have is we are doing something that's clearly making the world a better place.
Sal Daher: Yes, exactly.
Steve Shapiro: Yes. We're not destabilizing democracies or putting small businesses out of business. That really relates to the other topic I said I'd talk to you about, which is culture. I know most people are like, "Okay, I'm going to exit out of this podcast. I don't want to hear this boring-- I keep hearing culture, culture, culture from everybody," but how important culture is and that it can only be a competitive advantage.
The Right Approach to Company Culture
You have to be careful when you use the term culture because from a diversity and inclusion perspective, a lot of people's version of, "This is a great culture," might work. What I learned is the way that the term had been used in Silicon Valley over the last decade is actually quite toxic. When you’re thinking [crosstalk] from a diversity and inclusion perspective, massively toxic.
When I'm talking about culture, what I'm talking about is the culture of listening. The culture of the leadership team really being good listeners, the culture of everyone on the team really understanding what is the big picture of what we're doing here? What is my role, and as a company, what are we trying to do in the greater scheme of things here? Grounding people in that and making sure that you're continually updating in that...that changes. That changed for us.
We went from, "Hey, it's a manual human-enabled workflow, it really helps a teacher," to, "Whoa this thing is using the latest AI. Quite honestly, an AI that some people are writing about negatively, so we're going to use it positively, we're going put guardrails on it," but just grounding your team in that and helping them understand the vision. That requires continual re-calibration and reevaluation.
When we talk about culture, that's what we talk about. Sure, we say things like, "What's most important to us is mutual respect," so I want to say, of the 62 people on our team, one thing we can brag about, and everyone could say this, is there's not one a-hole on the team. How do you tell that to-- You could tell that to someone who's a candidate to join you, but they won't really be able to know it till they actually join and see, "Whoa, people have actually been taught how to respectfully disagree and work through problems in a professional way, and not taking stuff personally." If you can work that way, you can move mountains.
That's all part of the way we think about culture. We don't spend a lot of time blah, blah, blahing the usual stuff about it, we just think. It's mutual respect, it's everyone understanding the big picture. Obviously, for leadership, it's being a great listener and, and walking the walk instead of just talking the talk. I hope that didn't sound so vanilla?
How a Founder Learned to Get Along with Her Co-Founder
Sal Daher: No. I think what it points to is the possibility that the relations between human beings can be improved. Today, I was listening to an interview with Jessica McKellar, who's a founder of Pilot.com, and I don't know Jessica, but I know the other two co-founders that she had. It's funny, in the interview she says, "I've met Jeff Arnold." She says for seven years, they've been together for 15 years and they're on their third startup together. She said, "For seven years, Jeff and I didn't get along."
Steve Shapiro: That's a long time.
Sal Daher: Yes, that's 7 out of the 15 years, but along the way, she went through executive coaching program at Dropbox and it taught her some things that helped her understand the value of people who may have a different perspective, and also, not everything has to be exactly the way that you want it to be. Perhaps you think, "Oh, it's probably better this way, but this other person is very tied to this thing, let's figure a way to work it out." She didn't say the 15 years that she's been working with Jeff they couldn't get along, it's seven years. The other eight years, I guess, they have gotten along. It's a very hopeful thing.
Steve Shapiro: Hopefully, it's the latter eight years that they got along.
Sal Daher: Exactly, they got along really well, but the last seven years have been hell. It’s the other way around. I think it traces back to this purposeful effort to understand how to relate to your co-workers, how to relate to customers as well, which someone like Jessica McKellar, who's just incandescently bright, you'd expect someone like that to be able to make that shift, but I think-- Was it Henry James? One of the James' family said something that the difference between the most intelligent human being and the least intelligent is not that much. In order for us to function to do what we do, we have to be tremendously, tremendously capable. We are unbelievably capable learning machines.
I think that every human being has that capacity if they are put it in the situation where it can be helpful. I think this is where a company culture sounds very trite, but there's something in the Dropbox company culture that is really, really helpful, that changed the relationship of Jessica and Jeff.
Steve Shapiro: It sounds like Jessica and Jeff learned that old adage at end of year seven, which is, "Do you want to be right or do you want to be happy?"
[laughter]
Sal Daher: It could have been in year one if they had heard that, but it's not like that. The thing is that the context has to be that has to be presented at the time when you're ready to learn it.
Steve Shapiro: Yes.
Sal Daher: You could have said that 100 times before. I think back in my own life, until you're ready to learn the thing, you cannot learn it. This is why adaptive learning is so important, you have to meet the student, the learner where he or she is, you cannot expect that it's going to happen just magically.
Steve Shapiro: That's right.
How Younger Workers Are Disadvantaged by Remote Work
Sal Daher: Now, one of the things that strikes me about remote work is that if you are someone who has been, let's say, in a line of work for 10, 15 years, 20 years, working remotely with people that you know is not a problem, you don't lose much by it. If you're someone who's getting into the profession, it is a massive problem, you don't have the sort of social networking capital, so to speak.
There are little things that remote communication do not transmit very well. This is for the young, we have to think in terms of what happens to the young people in organizations? How do we bring them into the network? How do we socialize them into the culture of the company remotely? Do you have any thoughts on that?
Steve Shapiro: Yes, I'm glad you brought it up because that was actually my next topic I wanted to talk about. It's near and dear to me because I've two children in their 20s. Through our son who's in his late 20s and currently has a remote job, I think about this a lot, and through our employees that are in that.
One of the things is that when you think about, in my case, the people in my network that have had great careers, there's a couple of similar themes that most of them had, and that was that somewhere probably in their 20s or in no later than their 30s, they were in an organization where they established early great mentor or mentors. It doesn't need to be one person, it can be a number of different people that were either boss or someone senior to you in the organization that is your lifelong person you could call and something like that.
I worry for this generation of people in their 20s right now that especially getting started during when this pandemic started, that's so much harder. Then, what can we do as either an employer or just someone in the network to help them? I think the best thing that we can do is if you could almost normalize like it's almost part of my job to help you network and build a network. Even though right now because there's a pandemic, some of it is doing it the way you and I are communicating right now, but can I start to introduce you to some new people?
Sometimes it actually doesn't have to be so, gosh, darn transactional because the best networking is not transactional, it's not, "Hey, I want to meet you because I want something from you." The best networking is, "I want to meet you and help you." Really. Right?
Sal Daher: Yes.
Steve Shapiro: The advice that I give to a lot of people in their 20s are, and if I can help them and say, "What would you like to do? How would you like to get out there and meet different people, either in your field of interest or a field that you might be getting into? What are ways that you can meet them and potentially do something for them? It could be small, and could I help you do that? Maybe you need a little help."
If you happen to have someone in your network like myself and I can help facilitate it for you or I can at least give you the guardrails of, "This is what you should do. You should reach out to this guy," or "I'll connect you to this guy," or "Hey, what's your biggest problem right now?" Is there someone that could do some grunt work? I know people don't like to hear that, but for all of us, no matter where we got, when we were in our 20s, we did a lot of grunt work and that's part of the learning process.
Sal Daher: You're young and you are capable of learning new tasks, physically you have more stamina when you're younger, and it suits you for doing stuff like that, through which you will learn some of these skills as your cognitive abilities decline, as your physical abilities decline, you'll still be able to be productive because you've built a social network and social capital, so to speak.
Steve Shapiro: Exactly. On top of that, the moment that they say, "People are mostly going back to the office," hopefully, find a way back to an office. If you are in a situation like our company where we're probably not going back to an office because we're in 15 different countries, but what are the things you can do? We want people like, "Hey, do you have a WeWork or some co-working space where you live? I want to fund you being able to go there. I want you to be around people. Are there local association that you can join?"
There's something some way where if you are working out of your living room, there is something pulling you out during the week and there's a group of people and things like that. I think those are some of the best practices to ensure that if, for some reason, it is working for you to be working remote and the company is remote, what are the other things you could do to have a more full professional career?
Sal Daher: This is really good. Call to action to managers, think about not just how to get your workers to meet the goals that you have, the deadlines that you have and so forth, but spare a little time for helping, particularly your younger workers, to develop their social connections by paying for them to get together in a WeWork.
Maybe if three or four of them are working in the same city, they have something in common, they can work at a WeWork from time to time or something like that, or connecting in with people, and just remember this, because if you are in your 40s, in your 50s, in your 60s, you have so much social capital that someone who's starting out on a career, a person in 20s and 30s does not have, and so you have a responsibility to be mindful of how you can help those people. It will pay off not just for you feeling gratified that you're helping another person, but also, it will help your company have lower turnover. People will have greater satisfaction.
Steve Shapiro: No doubt about it.
How to Help Younger Workers Build Social Networks? A Question for Jacqueline Olds, M.D.
Sal Daher: Yes. This is going on in all sorts of things. I'm preparing to interview a friend who is a psychiatrist. She's a professor at Harvard Medical School, Jacqueline Olds, and she wrote a book called The Lonely American. I'm going to be asking her about this, "How do we do this? What ideas do you have for this?" I'm listening to her book, on Audible right now [crosstalk]--
Steve Shapiro: I'm in. Don't worry, tell her she's going to have very strong listenership for her episodes. [laughs]
Sal Daher: Okay. She is so amazing, [unintelligible 00:53:55] It's funny, she's a founder of a startup. It didn't succeed. We're going to discuss that, but she's a founder of a startup, grew out of her passion to help people overcome depression. I was an investor and that's how I got to know her. This is an amazing lady. She's going to be on the podcast. I'm going to pop a question to you, Steve. Now, we're talking about helping young people connect socially, become more socially connected.
How Do We Reskill Angel Investors to Help Build Life Science Startups?
I have found an amazing opportunity that I'm pursuing, and that is what I call angel-scale biotech, early-stage biotechnology companies that can be funded by angel investors, no more than $5 or $6 million, usually an academic founder, and usually really strong technology that's of interest to strategic players. These are not going to be billion-dollar companies, they might be $100-something million, $200 million companies, but if all you've raised to get there is $5 million, there's lots of scope for angel investors to make tons of money. How do I reskill angel investors that have grown up around investing in software startups, which has been most of what's been going on in the last--
Angel investing has grown under the shade of the giant sequoia of software. Every angel has been doing something related to, but right now there's an explosion of opportunities. Some of it is health-related, it's digital, but some of it is just wet-lab stuff that is becoming really accessible. It can be done with a few million and it can be snapped up of $150, $200. I'll give you examples. One company is SIAMAB which is of someone who's been in this podcast. There's another company that is someone who's going to be in the podcast that gave a very nice 10X return to investors in technology that was taken from a lab and is now part of a big strategic player.
How do I reskill people? How do I bring these angels in and get them to understand that it's possible? I'm not a biologist. I'm not a life scientist. Without being a life scientist, it's possible for you to learn the particular knowledge that you need to acquire about the startup and to apply your years and years of experience in business to help this fledgling business.
Steve Shapiro: It sounds to me like the perfect moment for you to create a short-term boot camp for angel investors. I am an angel, as you know. I've been side by side with you on a couple deals. Through my formative experiences at Launchpad, where you get to know another a 100, 120 people, there's not one persona, obviously, but there is a common strand of the people that I've met there and the other angel groups that I'm in. Some people invest by area of expertise, some people are generalists, some people are risk-averse, which sounds so counterintuitive to be an angel investor and risk-averse, but they're risk-averse to going into something-
Sal Daher: They don't want to be in the stock market right now.
Steve Shapiro: [laughs] They're risk-averse about going into something they can't understand. One of the best things that boot camp could do, and again, it's not going to give them an advanced degree in biology or anything, but is the idea of grounding them in some journeys that people have had in both successful and unsuccessful biotech investments, and helping people think about what is the mindset here? What are the possibilities?
What I would do, almost Harvard Business School case studies of two or three deals, maybe two that were successful, one that failed, and to ground them in the possibilities and possibly do it like Harvard Business School style. I'm going to present you with the case and we're going to work through this, like we're at HBS. I think that would be a real formative experience for a lot of angels and help get them past what some of the trepidations are when you don't understand the product per se, or you don't understand the problem it's solving well enough or how it's the solution.
Sal Daher: I like your idea of case studies of the investment, like a narrative about the investment in certain companies.
Steve Shapiro: Sign me up to help you on that.
Sal Daher: I would love to do that. See, I can think of one company that would be a company that failed, a company called Parabase Genomics. It gave rise to a company that is just killing it right now, which is Meenta. Have you run across Meenta, the- [crosstalk]
Steve Shapiro: No, I don't think so.
Sal Daher: They have a SaaS platform to connect people to testing resources, whether it's a COVID test or next-generation sequencing at the Broad Institute or something like that. It used to be that if you wanted to run an experiment on someone else's machine, it took 60 emails. What they've done is they've created a process where you can just go online to find the right machine, the right time, and book a slot on some multimillion-dollar machine, and do you experiment and get your results, and get the right reagents and all that sort of stuff.
The company they did before, one of the founders was a failure, and it astonished me that four months after closing down the company, he called me and said, "Hey, Sal, let's have some coffee and let's talk about my new company. I want to start a company," [laughs] and it is succeeding crazy right now.
Steve Shapiro: That's great.
Sal Daher: I think there are some really interesting stories to be told.
Steve Shapiro: Happy to help you on that.
Sal Daher: I'm going to take you up on this.
Steve Shapiro: We have 10,000 witnesses that will be listening to this podcast.
Sal Daher: Exactly. No, and I'm just thinking that perhaps you could be a guest host on interview with the founder talking about the experience, the failure, and then the success.
Steve Shapiro: Would love to.
Sal Daher: I think that would be really very super, but with the purpose, the stated purpose of helping angel investors understand what it is that goes on with these life science companies.
Steve Shapiro: It's the greatest idea. [crosstalk] Well, not the greatest, I think the greatest idea we have right here on Finetune, but it's up there. [laughs]
Sal Daher: Yes. I'm a very happy investor on Finetune Learning. [crosstalk]
Steve Shapiro: Thank you.
Sal Daher: You guys should find out why, listen to “Miracle-Working CEO” podcast, the first one with Steve [Shapiro]. Steve, is there anything else that's on your mind that you want to touch on?
Steve Shapiro: Yes, I just wanted to-- This is more of like a public service announcement, but I still feel like I want to do it. I think that the last two years have been horrible for most people, almost everybody. I hope that we can find a way out of this and learn from it. I think it's affected our most vulnerable members of our population. I think those of us who are less vulnerable have a duty to get out there and think about ways to help those who are less fortunate. This has just, I think, in some respects, accelerated my call to action for people who've been fortunate. That's my public service announcement.
Sal Daher: No, it echoes what we discussed before. Some of the people who are most unfortunate in this are younger people, people starting out in their careers. You would think that these are highly privileged people, these people who are going to college and all this, but their lives have been turned up-- I could not imagine if my college years I had been forced to be locked up in my room, not being able to go out for extended times. I would have gone crazy. I think we older people, more established people have to think about this. Steve, there is a ray of hope. I'll tell you something interesting.
Steve Shapiro: Please, tell me.
A Happy Outcome from Zoom School
Sal Daher: It’s about Zoom school; Zoom school is terrible in so many ways. I have a seven-year-old son—grandson.
Steve Shapiro: I was like, "Whoa, Sal, I did not know that."
Sal Daher: No. That was scary. That would be so much work. A seven-year-old grandson and a four-year-old granddaughter, and because they were doing Zoom school, they were home all the time, the relationship between the two of them developed tremendously to the point where they would play. Before they didn’t play with each other. The seven-year-old had his friends, the four-year-old had her friends, they would play together. They began to play together and they became great buddies. Doesn't mean they don't fight once in a while. [crosstalk]
Steve Shapiro: The unintended consequence of they had no other choice.
Sal Daher: Exactly. We ended on a hopeful note.
Steve Shapiro’s Parting Thoughts
Steve Shapiro: Yes. I also, if you don't mind, one other shout out. That's to everybody who's a teacher and involved in education, especially K-12 education. This last two years has been hell for them. It's been brutal. I don't think that society in general has acknowledged just what heroes they have been and how challenging the situation has been for them. It was actually really challenging for them before the pandemic, but hundred times worse when the pandemic came. As a society, we really have to rethink how we resource it, how we treat it because this is a great example of we're not doing it the right way right now.
Sal Daher: No. I cannot imagine what it must have been like for them to have to all of a sudden adapt to teaching kids over Zoom. Great. Well, Steve, I thank you for making the time and for making your very generous offer to help me with my biotech angel boot camp.
Steve Shapiro: Love to.
Sal Daher: I'm got to take you up on that. Continue to stay well. and thanks again.
Steve Shapiro: Always a pleasure, Sal. Thanks so much. This is one of the things I look forward to every, I don't know, three years.
[laughter]
Sal Daher: Steve Shapiro, miracle-working CEO of Finetune Learning, thanks for being on the Angel Invest Boston Podcast.
Steve Shapiro: Thank you, Sal. Really appreciate it.
Sal Daher: I'm Sal Daher, thanks for listening.
[music]
Sal Daher: I'm glad you were able to join us. Our engineer is Raul Rosa. Our theme was composed by John McKusick. Our graphic design is by Katharine Woodman-Maynard. Our host is coached by Grace Daher.