HLS in the World | Tech, Law, and Law Teaching

HLS in the World | Tech, Law, and Law Teaching

Articles, Blog , , , , , , , , , , , 0 Comments


CHRIS BAVITZ: I think we will
get started with our panel this morning. Thank you all for being here. Welcome to the Friday sessions
for the HLS Bicentennial. It’s a real honor for me to be
moderating this terrific panel here this morning. Not with– as Kendra was
pointing out, this kind of room seems to lend itself
to cold calling. We won’t do that. But there are as many people
in the room I know already. So we are seated
with people who have a great interest in this topic. So we’re going to
talk for a while. And then we’re going to open
things up for conversation with those of you in the room. We have till 10:30. And so again, we’ll do some
panel discussion here, and then be sure to draw you all in. This is a topic that
we’re talking about today that’s of great interest to me. It’s of significant interest
to folks at the law school. It’s been of interest to people
at the law school going back quite a while. You and I’ve been emailing
a bit about some classes that she taught here
quite while ago on topics related to technology and law. And we have Jack
Cushman on the panel, who currently teaches a
programming for lawyers class here today. And the basic gist
of this, I think, is to talk about what we want
to be teaching as law teachers to law students
about technology. Now that’s not technology
law, necessarily. We have, and have
always had lots of great classes here on
intellectual property, and privacy, and security,
and online speech, and those kinds of topics. Nor is it specifically about
sort of technology practice skills training. My role here at the law school
is to run our Cyber Law Clinic, where I work along with
Kendra, which is our Tech Law and Policy Clinical Program. And we give students
the opportunity to work with startups,
and computer programmers, and others, on legal issues that
they face, and that’s great. But this is really
about responding to the interests of students and
kind of getting under the hood. So this is about coding. This is about technology
literacy, that sort of thing. We have again a great
panel here today. Rather than sitting
here and reading bios, I’m going to give you a
very quick introduction. And then I’m going
to ask each of them to maybe say a few
words about who they are and the perspective they
bring to the conversation. To my left, Mitch Resnick,
from the MIT Media Lab, a friend and
regular collaborator with us in the clinic,
and the founder of the Scratch platform,
which some of you may know. For my eight-year-old,
Mitch is a hero, because he’s a big fan
of the Scratch platform, but it’s really the primary,
the preeminent online learn to code platform
for kids these days, a really extraordinary project. To Mitch’s left, Kendra Albert. Kendra is an HLS alum. They work with me in
the Cyber Law Clinic and spent the past
year in tech practice, out in Northern
California before we were able to lure them back and
join us to teach and practice in our clinical program. The next, Jack Cushman. Jack is a lecturer on
law here, as I mentioned. Has developed, and co-taught,
taught for the first time last spring, and will
co-teach again this spring, our computer programming
for lawyers class. He’s also based at
our Library Innovation Lab, which is a really
extraordinary place, that I hope he talks a
little bit, and is a Fellow at the Berkman
Klein Center for Internet and Society. And last, but by no means
least, Raj Goyle, HLS alum, former Kansas State
representative, lawyer, tech geek, Co-CEO of
Bodhala, which again, I hope he’ll tell us a
little bit about as well. Mitch, can I start with you? Mitch is our token non-lawyer
on this panel, I should say, which is great. RAJ GOYLE: You always
bring diversity, Chris. CHRIS BAVITZ: Maybe say a few
words about your work and sort of how you come to the topic
of teaching about technology, teaching about coding. MITCH RESNICK: Well, first
thanks for inviting me. And because I’m not
a lawyer, I will say at some time
in my background, at a time of great
indecision in my life, I did apply, and was accepted
into Harvard Law School. CHRIS BAVITZ: Really? MITCH RESNICK: But did not come. That’s a secret that
most people don’t know. CHRIS BAVITZ: Well, it’s fixed
on camera now for all eternity. So, yeah, terrific. MITCH RESNICK: As
Chris mentioned, I’m a professor at MIT Media
Lab and run this Scratch program language online community,
which we launched 10 years ago and has grown to an online
community of more than 20 million registered
members of kids around the world,
ages 8 to 16, where kids programs their own
interactive stories, games, and animations, and then
share their creations with one another in a very active
online community that gets– every day, there’s
more than 200,000 new projects that kids around
the world create, and an equal number of
comments, where they comment on each other’s projects. Through this, through
our work on Scratch, we interact a lot with the
Cyber Law Clinic on everything from the use of
copyright material in the online community,
trademark infringement, issues about child data
privacy, issues around what are our obligations of informing
authorities, or teachers, or parents about certain
activities that are happening. So there’s lots of issues,
legal issues, that come up. It’s been great to work
together with Chirs and others at at the Cyber Law Clinic. One thing I want to highlight
is, as we started Scratch, as we continue with
Scratch, our goal was not to prepare
the next generation of professional computer
programmers or computer scientists. Although some kids might
very well end up with that. That was definitely
not our goal. Our goal, our feeling is that
it’s valuable for everybody to learn to express
themselves creatively with new technologies,
and coding is a good way of doing that. So we see the goals at
Scratch to help people not just learn particular
technical skills, but to develop their thinking,
to think systematically and in new ways, to be able
to develop their voice, so they can express
their ideas in new ways, and to help their
identity, to feel that they can be an
active participant and contributor in a
society where technology is playing a very central role. So I think the same
way that we try to– truly, it’s a type of literacy. The same way we
don’t teach people to learn to write because
we expect them to grow up to become professional
writers, that’s our same goal with
8 to 16-year-olds. So it will be
interesting to see how our goals with that
community translates over into some of the
discussions around what happens in law school. CHRIS BAVITZ: That’s great. Thank you, Mitch. Kendra– KENDRA ALBERT: I feel like it’s
hard to follow up from someone who’s taught so many
folks how to maybe– if not to learn
to code, then how to be excited about the
practice of computer programming and technology more generally. But I sort of admit, I was a
little surprised when Chris asked me to be on this panel,
because my contributions, besides being a
technology lawyer, included one year of Java
in high school, which I think went quite well, but
I chose not to continue with. And then a sort of failed series
of attempts to learn Python before I started law school. And I think one of
the things that I think a lot about when I
think about the practice of technology, and both
technology law and more technology literacy
more generally is when I started
learning Python, before I went to law
school, one of the things I was looking for was a broader
sense of technology literacy. And I thought that
one of the ways to do that was by learning a
specific programming language. So I did perhaps all the
right things, perhaps all the wrong things. I joined the Boston
Python users group, which was very kind to me. And then I did project
Euler problems, which are these
little math problems to teach you to start
exploring different functions and language. And I discovered that I was
not learning any of the things that I wanted to learn. And so, I think
more broadly, often people turn to learning to code
and use the phrase learning to code, when what they
mean is, as Chris said, technological literacy,
sort of understanding the fundamental ways in
which technology works, and what technology
can do and can’t do. And I’m previewing
what Jack is probably going to say a little bit. But I think
technology literacy is important for all
kinds of lawyers, because it helps us evaluate
the claims that people make. It’s the equivalent of
being able to understand basic statistics to see whether
an expert is telling the truth or not. So I think that
technology literacy is incredibly important
for my practice, because I work on
computer security. So it’s literally
my job to sometimes be the go-between
between the lawyers who do not understand computers
and the client who does. But more broadly,
I think that when we ask people to evaluate
the claims of technology as it’s being used in courts,
in the workplace, in society more generally, that requires
a baseline understanding of how computers work and what they’re
currently able to accomplish. So if I were able to get
law students to think about these kinds of
things, and sometimes I try to, even as part of
my teaching practice, I would rather them be able
to understand that the word algorithm should be
more closely substituted with complicated
math than learn how to write a script in Python. So I think that good technology
literacy teaching teaches people to unpack concepts and
to more closely relate them to what they already understand. CHRIS BAVITZ:
That’s great, Jack– JACK CUSHMAN: I was
a little worried about where it was going
with that introduction because I do teach Python to
Harvard Law students as a way to teach technical literacy. KENDRA ALBERT: If had
taken it from you, I probably would
have stuck with it. JACK CUSHMAN: I hope so. I was starting to feel
a little guilty though, because it is the risk that
we have this idea that this will make an impact. And it might turn
out that it doesn’t. And we don’t have really
many models to work from. Our course is inspired
by another course that was taught at Georgetown
by Paula and Jonathan Frenkel for one year. And we looked at them, and said
that’s great, let’s try it. And now we’re trying it. We think it’s making
an impact, but we might look back, and say,
sorry, Kendra, that was wrong. But here’s why we think
it’s making an impact. For me, the term
technical literacy is almost too
small for the thing that we’re struggling
with right now. I sometimes say
computational thinking, though that also feels small. The core thing
that’s happening is that our world is being
rewritten by technology, and lawyers have to help
us make sense of that. I saw that as first, a
computer programmer, and then as a lawyer myself, practicing
appellate law in Boston. Of course, being a
computer programmer, and then working as a lawyer,
I kept seeing opportunities. Oh, I could script that up,
and we could save that person 20 hours if I write this thing. We can win this case
if I write this thing. But that wasn’t the
biggest part of it. The biggest part is that when
you practice appellate law, you have to win by
explaining to the judge where the law should go. You tell them a story about
if you make this decision, the law will move
in a good direction, and you will have done your
work as Supreme Court judge. And when you start thinking
about where the world is going, you see that every
bit of the world is being fundamentally
altered by the technology that we now have,
and altered fast. I was reading, in
preparation for this– how is document
review changing, how is the demand for first year
and law associates changing. And you read these things, like
the practice of document review has fundamentally changed
in the past five years. And I wonder– each
of us could pick our– what changed for me since
2012, or since 2007. But if you think about
it changing fundamentally in the past five years, that’s
not a sort of legal practice requires you to learn
the new stuff problem. That’s a people who are
graduating from law school now when they made their
decision to go to law school, the practice has fundamentally
changed since then. Someone who asks
you now should I go to law school,
the right answer is, I only know of the
legal profession in 2017. I don’t know it in 2021. I don’t know if you
should go to law school. And that pace of change
requires this thing that we call technical literacy,
computational thinking, the ability to see how
it is that the world is going to be rewritten. Not only because document
review is going to change, but because we rely on
lawyers to make the world work after technology breaks it. And when we’re
training lawyers, they have to be able to then
go out and do that. So that’s what I hope
my class is able to do, and maybe we’ll get into how
it does that as we go along. CHRIS BAVITZ: That’s great. RAJ GOYLE: Good morning. That could not have been a
better segue, actually, Jack. I’m honored to be here. And I gave a talk at Berkman
that Chris had me last year, and I’m shocked you
invited me back. But thank you, Chris–
for many reasons. But I tend to find myself to
be a bit of an outlier on most of these panels and a
fairly strong critic of this institution. And I think I will
say some things that I don’t think any public
speaker will say about what’s actually happening
in our profession, and what a failure our industry
and profession really is. But I’ll get to that later. I’d love to– and then we can
talk about Kansas politics any time anyone wants to. Because I have
this logical thread of being a state rep
out of Wichita, and now I live in Tribeca and
run a tech company. So you know, I like
non-linear progressions. But to Jack’s point,
which is exactly right, which is I would not
counsel, and am now, because I do have an extensive
political background, about advising
young folks who want to go into law school and so
on, and this default degree that we sort of– I’m 42. And so, from 25 years
back, gosh, you know, when I was making
these decisions, you know, I don’t think
I could counsel anyone in good conscience to
actually say, oh, yeah, law school is a great place
to go, it’s a general thing, figure it out. And I will tell you, and I
say this from the front lines, it is great, obviously, you
know academics have of a role. My bias is toward doers. And so I actually run
a technology company that is on the leading
edge of doing everything that Jack was talking about. So we are transforming
the practice and the industry of law. So the discussion, in my view,
and I’m going to say we a lot, and this is my company,
and my co-founder, who is also an HLS
graduate of ’99, we sort of think about
this in a similar way. Which is, so should
the discussion be, should lawyers and the
law do coding and so on? I think, absolutely. I mean, it’s sort of
like a no-brainer. Of course, all of this
great stuff should happen. I mean, would anybody reasonably
argue that the law should not get more facile in technology? You’d have to be a complete
moron to think that. But the real, I think, the
bigger question, again, coming after Jack, is that
the AI revolution, of which we are one leading edge in
this industry, and the law has been hugely backwards
when it comes to technology– I mean, laughably
backwards, which is why there is whitespace for
a company like ours to develop– is actually missing, in
our view, the main point. Which is it’s not really
about these policies and what the
pedagogies should be, it’s actually about
making markets. And again, as Jack
was mentioning, the market is changing
very, very rapidly, because the law for
50 years has been able to survive in
its current form, due to a complete lack of
any market transparency. So only in monopolistic
markets typically do you have the supply
side setting prices. But in our industry– I loved when you
said, oh I could write a script to save 20 hours. You’d be fired
from that law firm. Why the hell would
a law firm want you to save hours billable? It’s like to anyone who’s in the
real world of this, it’s like– it’s not even– I mean, it’s like arguing
night is day, and day is night. That is absolutely
antithetical to the nature of the institutions
that actually run the legal industry. And so you have to think
with that frame really. And so our view is that the
way that intelligence software and AI will actually affect
our industry and the profession will be that it’s going to
change the market very rapidly. And we’re seeing this. We are a software provider
for some of the largest corporations in the world. And now some of the largest
law firms in the world. And the law has lacked what– actually McKinsey, the head of
the McKinsey legal practice, calls an OS. There is no operating
system in the law. It’s essentially a
bunch of siloed areas. And they talk to each other
a little bit, but everybody– you know, it is not
actually wholly– a tapestry like it is
in other industries. So I would just conclude
this little point with an analogy, which is if
you look at the way pricing transparency has changed
other industries, it’s really not about robots
eating jobs, or so on. If you look at Google and
Facebook, what have they done? They’ve destroyed the
advertising market. So it’s a market force,
where they came in, and essentially if you’re
doing any advertising, within a really
rapid amount of time, they have completely
destroyed the existing status quo of advertising. That’s actually how you get
transformational change. It’s not necessarily because– again, that’s the power of
the private sector, verse– and having an extensive
background in the public sector and believe very
passionately in it, you know our bias, from
our view of the world, is that you have to
watch these market forces far more closely
than is actually being done in the academy. CHRIS BAVITZ: So much there. Lots of threads I’m going to
try to start to weave together. Those were great intros. One of the things I think
I heard from all of you was sort of this trying to draw
this line between understanding how technology works,
understanding how code works specifically for the purpose
of writing a program, or writing a script, or
generating a privilege log for a lawyer, or
creating an iPhone app for a general
programmer, on one hand. And understanding
how coding works for purposes of gaining
general computational thinking skills, technology
literacy, tech fluency. And I’m curious, Mitch,
if you have thoughts, on the earlier stage of teaching
kids to engage with the stuff, you started by saying
not all of these kids are going to grow up to
become computer programmers. But can we drill
down a little bit on exactly what we think kids
are learning when they’re learning coding skills, if not
the logical, practical result of coding skills, which is
the ability to write software. MITCH RESNICK:
Often times, people start by saying you get
a type of systematicity in your thinking. And I think that’s true. But it doesn’t go far enough. Actually there was
some comparison that was made about learning math. And is this the equivalence
of learning math. And although there’s an
equivalence in some ways, I think there’s an
important difference. In fact, one thing was
really influential to me, when I was first
getting involved in doing things in computer
science, I read this book. I’ve always just seeing computer
programming as something just to get a job done. It was very functional,
my view of it. I learned programming,
get a job done. And then there was this
computer science textbook from MIT that said
computer programs should be written primarily for people
to read and not for machines to execute. And I said, what’s that mean? You know, I thought
that’s shocking. I thought it was this
operational thing to get a job done. And what they meant by that is,
it’s a way of expressing ideas. And it’s a different
way of expressing ideas than mathematics is. Mathematics generally expresses
declarative knowledge. There’s a fact, this is
true, there’s equality. Computer programs
describe process. So one reason I think learning
programming is valuable is it gets you
thinking about process and gives you a
description of process. It’s a new way of
describing process. And we’ve always been very
bad at describing process, and thinking about process. And actually, in
the field of law, it’s important to
think about process. So I’m not saying that everyone
should write a computer program to describe the
processes that you’re studying, but it gets you thinking
about process in a way that I think few
other things do, I think in a very valuable way. CHRIS BAVITZ: Jack, or
Kendra, does that resonate with either of the way of your
teaching skills, or Kendra, the way you’ve thought
about them in your practice? KENDRA ALBERT: I
think actually if I were going to be
like why I should– if I were trying to come up with
the most compelling reason why students should
take Jack’s class. Well, maybe not Jack’s
class, but a class in computer programming, I
would actually think that– one thing I always think
about is how many other places this conversation is happening. How many other people can sit
up here, and I’m pretty sure that’s actually a panel later on
how the law should be learning statistics, or
how the law should be learning more about all
of these different subjects. And so I think one of
the most valuable things that students can get out
of learning programming is not just the
systematized thinking, but learning to
deal with failure and learning to
deal with the idea that they don’t know
everything, and they’re not going to know everything. And I think that actually– I don’t mean– I’m going to join Raj’s down on
law school train a little bit, despite teaching
at a law school, in the sense that
I actually think we’re really bad at teaching
law students how to fail and how to find
things really hard and have that not be a sign
that you’re doing it wrong. MITCH RESNICK: If I
can just chime in, I totally agree it’s
important to try things that are challenging. But beyond that, another
thing that I find really about programming is it’s a good
way you learn about debugging. Compared to a lot
of other things, you can see where
something works, and you can incrementally
try to fix it. So it’s not just helping you
see that you’re necessarily good at something, and
experience failure, but also learn how to deal
with failure and incrementally, and iteratively improve it. And of course, you do
that in everything you do. You write a paper, you
incrementally improved it. But programming tends to be
a particularly good context for learning about the process
of iteratively refining things, which I think is so
important in so many things, and how to deal with
failure by refining and adapting to the
way you deal with it. JACK CUSHMAN: Could
I just follow up? One other thought, programming
teaches you the process. Whatever it is you’re
studying, programming makes you say very clearly
how the process works. From a legal
teaching perspective, that has a huge opportunity. If you want to teach something
like systems thinking, if you want to teach– let’s implement this
very practical thing, like how does someone move
through the immigration statutes. And we will then have to
think very clearly about how that statute works
on a functional level, but that also opens up
great teaching opportunities for talking about how that
statute works in a systems thinking level,
on a social level, what changes would we recommend. And then we can test
those with a program. So it’s not just now
you’re better at thinking about a practical process. But if you have the right
sort of teaching in place, then you can say,
now let’s learn to have all kinds of other
process thinking that will make you more effective. CHRIS BAVITZ: Raj, because
both you and Jack talked in the introduction about the
rapid change of the practice, which is true for– I think everyone in the room
who practices law has seen this. How do you think we should
be preparing students to be using technologies
like Bodhala, like other legal
technologies out there, that use artificial intelligence,
that to Jack’s point, assist with discovery,
that sort of thing. Is training them to use
those platforms enough? Or is giving them a look
again under the hood essential to having them think
critically about these things? RAJ GOYLE: It’s a great point. So you are probably predicting
from my opening comments, yes, of course,
you should do that, because I think that what’s
happening in the law– so let’s take the document review point. So what happened
there, and I know we have a distribution
of ages in the room. So I think there’s
probably somebody here who started their career
in a dusty warehouse looking through banker’s
boxes, doing what’s called due diligence, which
is really just a price arbitrage from a older
person at that law firm who was making tons of money
off that human being doing work that was really
mind numbingly– I mean, hats off if you
could have done that. I would have jumped off
the roof of that building after about two
hours of doing that. So what happened
is, we call that AI. It’s really, really not
sophisticated technology at all, which is a world,
again in the real world, called e-discovery. So that’s really
keyword searching. And so that happened
about 20 years ago. And what did that do? It really did
transform litigation, because no longer did
the person at Paul Weiss just send an army of 20 people
down to the basement of Aetna when they were doing
some massive litigation. So now you have some
computers doing that. Did that change
litigation staffing? Sure. It didn’t destroy the
litigation model, by any means. So now the discussion again
on the front lines of this, when we’re dealing with
the banks and the insurance companies, is AI going to
do contract management? And I think you alluded to that. Well, sure it will. Of course, it should. Because you know, banks
are hardly rational actors, and they’re hardly efficient. However, over time,
they’re directionally perhaps efficient. And so they will use
efficiency tools. But what we disagree
with is the notion that somehow AI is going
to eat all the lawyer jobs. So that we think is a
really facile observation. I mean, I don’t think
there’s any evidence that suggests that. What really I think
you find in industry after industry, which is– remember email was supposedly
going to kill a ton of jobs. Well guess what, it did
kill a chunk of jobs. And then it created
tons of others, a ton of other kinds of jobs. So we’re in the
really early innings of where this is
going to go and how it’s going to affect the law. I think, consistent with that
other question you asked, Chris, which is I think
the problem that we see is that legal education,
with the exception of this panel, which is
incredibly interesting to me. And God, I wish I had had
this kind of discussion when I was here,
instead of third century comparative Bulgarian
law, which there was an endowed professorship for. And I think the last thing– I think somebody told me the
last innovation in the law was the Magna Carta, and
they’re probably being generous. I mean, think about legal
education, 204 law schools. Do we need that
many law schools? Do we really need 40,000
lawyers coming out every year. And by the way, 203 of them
try to be this institution, the Langdellian method. This is something we
promote, as if all law schools should be
autonomous robots, parodying another school. We should be innovative. Every single one of those places
should be a node of innovation, and interest, and
locally rooted. And so that’s why I think the
reason I’m not bullish on where this is going for the law,
is because the law has proven itself, data doesn’t
lie, decade after decade of resisting innovation. And in this case, the data
revolution is happening. And so the lawyers,
and the law students, and the legal establishment
should not bury its head in the sand like it
has typically done at every step of the curve. In fact, there are great stories
where lawyers resisted email in the 1990s. They said, my clients
will not permit this, their sensitive
documents in electronic format. They need to be in pouches,
you need a courier them from firm to the courthouse. So obviously, the question,
it’s still pending, all of these things, in terms
of where it’s going to go. I would just say
our view is that we are moving into a world of
what we call value pricing. So the business of the law for
50 years was the following. It was every law
firm cranks units, and those are billable
hours, and every year that unit price goes up. And so what’s your business
model, crank more units. That was my point about– so it was literally against
the business model of a firm to do what you were suggesting,
even though it would logically make sense. In the halls of Harvard
Law School, you’d do that. At Ropes and Gray,
you don’t do that. Now we’re moving into the
world of value pricing. So corporations will
no longer just accept the price of a legal service. And so now you have this
50-year sclerotic industry trying to sort itself
out of, oh, my gosh, I actually have to use data,
AI, market making software to determine my value. And so that’s a
long way of saying we don’t know how this is
going to actually affect legal education and the law. But it’s coming,
and so therefore, again to your great print,
Jack, which is in five years, the chairman of Paul Weiss– and
then I’ll stop here, I swear– which is that
Chair of Paul Weiss said that the actual
practice of the law bears no relation
to five years ago. And think about what that means. Paul Weiss is a multi-billion
dollar law firm. So the world is changing
so rapidly out there, again on the front
lines of this stuff, we just don’t know
where it will go. KENDRA ALBERT: So
I think I’ll take a weird role of standing up
for, what did you say it was? Third century comparative
Bulgarian law studies? RAJ GOYLE: That’s what
I took in my third year. Second year, I was more
focused on Estonia. KENDRA ALBERT: In my
time, we were already in 15th century comparative
Bulgarian law studies. But I think, you know– Raj, I think that your
experience with these law firms is like super valuable. But of course, not
all of legal practice is confined to large,
very fancy law firms. And although we may think
of much of the law that way, I think that– I was actually thinking of
and preparing for this panel, and I think that what changed
my experience of the law and my ability to practice
technology law much more than probably
my one year of Java, which I will reference
in every comment, just to prove my
chops to be here, is actually the work that I
did in critical legal studies, my only classes
without Duncan Kennedy, because they prepared to
better understand and unpack the role of systems in
cabining our thinking. And I think that good
programming classes and good computer
science classes can very much teach
you those skills. So I don’t want to
only say that Duncan Kennedy or the
critical legal movement is the only people
who are able to. But I think more
broadly, this emphasis on this specific set
of skills, no matter how much it will revolutionize
the profession, which I actually do
agree that it will, for the vast majority
of it, is a result of the ways in which
technology has in some sense considered a higher
status, like higher value set of skills than law. And that’s not actually
something layers encounter very often. I think occasionally, when I
am feeling down on my practice, that I went to law school
so that I could finally tell technologists what to do
and they would have to listen. CHRIS BAVITZ: Good
luck with that. KENDRA ALBERT: Which they don’t. I am lucky to have clients
that don’t listen to me. It makes my life
more interesting. But I do think that
more broadly, we can understand the push towards
law schools thinking about like should lawyers
learn to coach, should lawyers learn about technology. As a change in
relationship, where law is used to being
on top, lawyers are used to being
the ones in charge. And now, actually,
their profession, their daily practice, and
also the world around them is getting
revolutionized by forces that are not theirs to control. And I think that our
way of understanding that should actually
not just involve learning about technology,
although I think that that’s useful, but also learning about
all of the other critical tools that we have to unpack how
those assumptions work. So I were trying to
tell someone about what would become them a good public
interest technology lawyer right now, I might say– I probably wouldn’t say skip
Jack’s programming class. But I might say skip the CS50. Take a critical
race theory class. Because you will
learn more that will be relevant to the what you
do on an everyday basis, in terms of actually predicting
the kinds of changes that are going to happen to society
as a result of technology, than from the sort of everyday
practice of even learning to write an app. There’s my controversial
statement of the panel. CHRIS BAVITZ: That’s great. It actually dove tales a
little bit with something I’ve been thinking
about, having Mitch here, who’s teaching
kids, and the rest of us who I think are
teaching young professionals and professionals to be,
which is that I sometimes feel like we’re in– I wonder if we’re
in an interesting, but sort of short
term, sort of time gap here, where we now
have students coming into HLS with significantly
more tech literacy and computer programming skills, certainly
than when I went to law school 20 years ago. But that when my 8-year-old
who loves Scratch– God forbid he goes to
law school some day– he’s going into law school
with much, much, much more tech literacy than your average
Harvard Law School 1L today. And so I wonder is
what we’re trying to do with courses like
Jack’s and with some of these other interventions in
the law school curriculum, sort of bide our time,
until the young kids that Mitch is teaching kind
of catch up to the profession. Or maybe another way
to ask it, is there a difference in thinking
about doing this for kids versus doing this for adults. And is that gap someday
just going to go away. I don’t know if
that makes sense. MITCH RESNICK: I was
thinking about that, because I was thinking that if
we do our job well, out then 10 years from now, all the kids,
all students entering Harvard Law School will have experience
doing at least some experience with programming. And I do think that to
have some experience is what’s most important, as
opposed to great expertise with. You have some
sense– some of it is to understand what’s possible,
what’s novel, what’s unusual, what’s not. To be able to have
that sensibility, you can get from just having
some experience with it. And I think that
increasingly, at least I hope, it’s not for sure, but
if we’re successful, then most kids will have
that experience growing up. Then one can look at how to then
enhance that and build upon it. It does feel like one thing
that’s coming out of this panel is there’s these different
threads of different reasons for doing it. For some, it’s just
having some sensibilities about certain ways of thinking. Some is having
some sensibilities about the way technology
is used in the world. And those are pretty different. And probably, and you’re
going get different things from different levels of
engagement with programming and other things. JACK CUSHMAN: Can we– CHRIS BAVITZ: Please, yeah. JACK CUSHMAN: Can we draw
out another reason to do it? CHRIS BAVITZ: Sure. JACK CUSHMAN: I
think it’s related to both of these points. To Kendra’s point about
the way that technology confers power in
our world right now. Our class is, it’s a specific
requirement you never have taken a programming
class before. Which helps us
practice when you’re teaching, because
people are working with the same background. But it also means
that we’re targeting people who might
otherwise decline to claim expertise when
they become lawyers and go out in the world. And both I and my
co-teacher, Jody Winestock, have had the experience of
being lawyers, and being young lawyers with
technical expertise, and therefore having the
balance of power shift in a legal conference
room, where we have partners, who have
been there for 40 years, and the tech concept
will come up, and all the heads
will turn to us. And all of a sudden, we are
being listened to in a way that we wouldn’t be
on any other issue. And that lets you do things
that matter when you graduate. It lets you kind of claim
your place in the world. MITCH RESNICK: Same thing
happens in a fourth grade classroom where
the students have more power than the teacher. JACK CUSHMAN: And
so when I think about is our course remedial? Is our course going
to go away when, if you succeed beyond
your wildest dreams? Then, yes, I think
that could happen. But I suspect for a long
time, maybe forever, there will be people who hit
law school still unwilling to claim that expertise,
even when they could, even when they do have that
experience with that technology that they could
claim expertise in, but they don’t know
how to have that voice. And I think it’s
another, it ought to be another goal of
programming education in a legal context, to
say go out there and grab the power that you
can claim when you are willing to speak up that way. MITCH RESNICK: Actually,
it does relate. When I said earlier,
we sometimes think about our work of help
people develop their thinking, develop their voice,
develop their identity. And we say develop identify,
it overlaps with what you’re saying about power somewhat. In fact, we draw that from
great literacy movements. Like when Paulo Freire in
Brazil and literacy movements. It wasn’t just to
help people get jobs, it was for them to develop a
voice, to speak up in society, and for them to feel they
had some power in society. So I do think it’s
very related to that. JACK CUSHMAN: And just to pull
on the critical theory thread a little more. The participation,
for example, of women in computer science
programs, we know is going down instead of up. And you may remember the
statistic better than I do. Is it something
like 7% now, women in computer science programs? Something that is startling
bad when you learn it. And so I think the view that,
oh, now all the kids are doing this, and everyone is
going to come up with it, is not reflecting this
sort of differential in who is willing
to claim it or not claim it in practice right now. MITCH RESNICK: 45% of
the Scratch community is young women. CHRIS BAVITZ: That’s amazing. MITCH RESNICK: But it’s
active effort to do that. So it’s an ongoing
effort to try to– RAJ GOYLE: If I could jump in
on that, which is– and Kendra, to your point,
obviously, it’s funny. Because when you say
lawyers typically think of themselves as being
on top, I have no doubt lawyers do think that. In corporate world, nobody
thinks the lawyer’s on top. The general counsel’s
office is not– it can be a power center. But it is not ever obviously
a revenue producer. It is not even C-suite. And so it’s interesting,
this is again the distinction between
what our silo of the world thinks and this is again the
trends that are happening, which is that’s why some
general accounts now want to call themselves
CLOs, Chief Legal Officers, because they want
to be thought of as C-suite. So lawyers, think of– we
think of ourselves in this way. But actually, when you’re
in the board at a company, it could be a startup
board, or whatever, lawyers are ancillary. They are a necessary function. But they are not
critical in a sense to the actual management
of the business. And so, I think it’s an
important distinction to make. And of course, it
would make sense. At the B-school they think that
CFOs, and CEOs run the world, and Wharton’s a big
consultant school. They think that way. This discussion about power
though is interesting. Because if you step
back and think about– I said in my very
opening remarks I think the law has sort
of badly failed society. So my line is always that
if Alexander Hamilton– who is all the rage now, thanks
to Lin-Maneul and Miranda– came back, he would
be aghast at what actually happened to this
gorgeous thing called the law. You know, the law,
when it’s done well, is what separates
us from the animals. I mean, you know, anybody
can create an economy, and you can have trade, and you
can have GDP and all of that. But the law is what
separates us from– we literally say,
lawless nations. I mean, my parents
come from India. You know, and they were at
a newly independent India in the 1940s. And they eventually left. They came to America,
and not just because they wanted to succeed, which
they obviously did, but because the law represented,
America represented this place where the law worked. My dad, who’s a very
liberal Democrat, but he will say almost these
sort of these very patriotic things, which is great. Which is like, anybody
can make it here, which is why I came
here, and I knew in India I couldn’t make it the
way that I could make it here. Is that not a TV– you can make it anywhere. But the point is the law,
because it’s a little bit like the frog being boiled. We’re all told because
we’re in this institution, these are the way
that things should be, these are the way things are. But if you step back
and think, I always have this kind of six words,
which is like poor sucks middle scared, rich obnoxious. What do I mean by that? The poor get no legal services. And I don’t want to
hear about pro bono. I know it intimately well. We work for Amwell 100 firms. The need of society, and I’m
a former civil rights lawyer, I’m a former local elected
official, I’ve worked with– I ran a nonprofit foundation. We gave to community groups. The need in society for
legal services compared to the service
delivery is so vast, that it will make you cry if
you really think about it. Go to a legal service
office in Wichita, Kansas. They are still on
rotary phones and paper, while we’re sitting here,
you know while we’re here. The middle class are
generally afraid of lawyers. That means a divorce,
that means an accident. And then the rich get
into legal services. My wife’s rack
rate at Proskauer, she’s a partner there,
is $1,125 an hour. And now, thanks to Bodhala most
companies no longer pay that. And frankly, thanks
to Bodhala now law firms are understanding
our price, and so on. That’s a separate discussion. But we should, we
really should– and this is where I’m a
bit of a scold whenever I speak at a law school– which
is, it’s incumbent upon us, you know, now there’s a great
awakening on sexual harassment, and there’s this notion
of being complicit. And we are all
complicit in a sense that this legal system is
really failing our society. And we should all– it’s not to
say shame on me, shame on you. But we need to think
about these things. And I do think that
technology is a great leveler. That was the great
discussion you guys were just having, which is why
technology can be so great. Like, what you’ve built
with Scratch is remarkable. And that’s an empowering thing. And so I always stress
to point this out, because I don’t think this is
ever discussed on a law school campus, which is just the
massive asymmetry between what happens here, and
the need, and I supposed claims of what the
legal profession actually does. And the reason why
it happens is because of the market forces
that move the axes of the law away from its roots. So I’ll just say this, William
Rehnquist in 1986 wrote a Law Review article in which
he decried the rise of law firms that have– wait for it– 300 lawyers, and billing
requirements of 2,000 hours a year. My wife billed 3,300 when she
was a third year at Debevoise, and that’s why she’s now
a partner at a law firm. And the reason he said that
was not because it was immoral, he said the ethics
constraints are too great, because when the market
pressures are so severe, the ethics actually can’t
stand up for legal ethics that are rooted in the profession. So because of the
market pressures, we’ve actually had a system
that’s gone a little bit awry. I do think that technology
can be an empowering point to get the law back to its
roots, which is when it’s done well, and it’s its
best self, is one of the most remarkable systems
that humans have ever created. CHRIS BAVITZ: Great points. We talked about a
couple of different ways that we think about
educating students and others about
technical topics, and coding in particular. And there’s obviously a
distinction here between sort of general purpose
intro to coding courses, either for kids, like
Scratch, or Kendra eluded to the CS50 course, which
is a very well-known course. It’s taught at the college here,
that a lot of students from, frankly, around the world, but
certainly around the university can go take. And I remember when Jack
was getting his class up and running, we had some
conversations about this, about what is the distinction
between doing something like Scratch, going
to an online learn how to program in Python class,
cross registering in CS50, and then taking a
dedicated class that looks at technology as
applied to something particularly specific,
which in this case is law. If you’re a grad student
at the business school, you could imagine a technology
for business students course. If you’re a grad student at
the grad school of education, you can imagine a technology
for education students course. So Jack, or anyone else, do
want to talk a little bit about that, about what you
think is beneficial about sort of the very targeted approach
that you’re taking, versus the sort of the broad approach? JACK CUSHMAN: So
I think the power that we’re seeing emerge as
we put together last year’s class was the power
of the case method, the sort of familiar law
school approach of you take real situations and debate
them, and learn from them, and then build an idea of how
the world works from that. So I’ll highlight
a couple of hypos that we used in the class. I mentioned the
immigration hypo. And what we do is we go through
an introductory Python language textbook. And each chapter, we
learn the chapter. And it has some hypo
that comes from it. Our hypo for branching
statements in programming involved immigration. And it involved a 2013 bill that
was proposed, and ultimately failed, that offered
a path to citizenship. If you go online, you can find
all these complex flow charts of how people were supposed to
follow the path to citizenship from start to end. And what we did, is we
generated 1,000 anonymous people in a spreadsheet, and we said
you represent a nonprofit who has a constituency
of immigrants who would like to take advantage of
a possible path to citizenship. You want to understand
how this bill would affect your constituents. So write a program that
will look at these people and figure out how many
of them would successfully be able to follow
this path that’s proposed in this statute. And then, for some
of our students who had an easier
time with that, we said take the next step, and
start modifying the proposed bill, and find the smallest
modification you can that will have the biggest
impact for your constituents. So, oh, if we just change this
from a two-year requirement to a one-year requirement,
then 20% of our constituents would then be eligible,
but this won’t look like a large
request in terms of the lobbying for the bill. So we’re kind of using
computation to understand how people think about lobbying. And if you think about– RAJ GOYLE: Sort of like a model. JACK CUSHMAN: As a model, right. RAJ GOYLE: It’s kind
of the way you’d– JACK CUSHMAN: But it links
to something very real, because it’s not like
we invented the statue to highlight this
programming exercise. It really was applicable. And then we could
kind of get into– students start
proposing, oh, I think this would be a
small change, I think that would be a small change. Then you can say, well, OK,
so why was that set there. Why does that farm
exception exist. Actually, there’s this
powerful lobby that will only make this possible. And what you thought was a small
change is really a large one. So as you explore the
flowchart with your program, you start to see that there’s
a social structure around it. And it still links to law
and computer programming. I do want to get
to the second hypo. RAJ GOYLE: Actually, I
do want to say though, let me put my former politician
hat on, that’s a game changer. So if you walked into my
office as a legislator, and instead of just having
usually some talking points that say you should
vote for HB34 or whatever, and you had a model,
and I could say, oh, if you tweaked this
amendment, this is the actual dynamic time
effect, that’s amazing. I mean, you know, I
was criticizing law. Wait till you get to a
legislature on being backwards. Let’s just say
information does not flow dynamically on the floor
of a house of representatives. JACK CUSHMAN: And when
we’re talking about power, bear in mind that tax laws
are already constructed this way for some people,
health insurance laws are already constructed
this way for the people who have those resources. Whole has access to
this kind of analysis and who doesn’t shapes
what the law does. RAJ GOYLE: Yeah. JACK CUSHMAN: But I know– Let me get in one more,
because I’m so proud of this. [LAUGHTER] The next week, we
had a new hypo, and we wanted to analyze on
the computer programming side how do you cross reference data,
and put data in dictionaries. And the hypo we came up with
was you work for an immigration enforcement agency. And you’ve come in possession
of an anonymous data set of 1,000 people who are
undocumented immigrants, which of course
was our spreadsheet from the previous hypo. Your task this week
is to correlate it with an FBI database and
de-anonymize those people, so you can enforce
the immigration. And the hope of that series
is to say to these things that we write, when we
look at data and programs, and things that feel
abstract, if you switch the frame a little
bit, all of the sudden it’s terrifying what
power they have. And although I think our
immigration system is deplorable, that wasn’t the
point I was trying to make, that our immigration
system is deplorable. The point I was
trying to make is whether you like
this or not, you have to respect the power of it. And my sort of best
self, when I walk out, thinking I hope
that that class went well, I hope that 100
times as those people go through their careers, they see
a pile of data and say wait, I’ve seen this before. I have to think so
much more carefully about what this means in
the computational context that we’re in now. KENDRA ALBERT: And
that’s why I tell people they should take Jack’s class. RAJ GOYLE: I’m signing
up, that’s awesome. JACK CUSHMAN: This could be
the year to learn Python. [LAUGHTER] CHRIS BAVITZ: Mitch,
you wanted to jump in. MITCH RESNICK: Well
actually, a few things. I also really liked
those examples. I think that does highlight
several things we’ve already talked about, because I think
in going through those examples, people, first, are
getting some understanding of systems and processes. Also they’re understanding
the power of being able to tinker with them. Because again, one
thing that I think is special about certain
things computationally, you’re supplying
them the opportunity to try things out, play what-if
games in ways you oftentimes can’t in other things as
easily, which can be compelling. But then you can also
see the potential dangers of people using that in
ways that– there are always implicit assumptions
in a lot of the things that are built into the models. So the challenge
of models, models are powerful for
convincing a legislator, but also can be deceptive. I was reminded, years ago,
there was this article, it was called Seductions of Sin. It was by Paul Starr, who was
a sociologist, first here, at Harvard, then at Princeton. And he was in the early
days of the Bill Clinton administration. He was a health care
economist helping out. He said that as he
was in these meetings, something was triggering,
that he was recognizing, some deja vu. And he realized he was noticing
the similarities between what his kids were using,
Sim City, at home, and then he was seeing these
similar types of simulations. But it then struck him, he
knew that it in Sim City, there were these
biases built into, it was like always good
to decrease tax rates. The kids were
learning in Sim City the value of tax rates, because
it was built into the model. RAJ GOYLE: That was
the Reagan Revolution. MITCH RESNICK: It was
still stuck in Sim City. So he sort became aware of– but the things he was doing in
Washington and the health care, people were just assuming that
this real because a computer generated it. RAJ GOYLE: Fascinating. MITCH RESNICK: So being able
to have students understand the seductions– both the
power and the seduction, and the potential problematics
of seductions of models. KENDRA ALBERT:
Yeah, I have– there is this quote I really like,
which is “All models are wrong, but some are useful.” RAJ GOYLE: That’s good. KENDRA ALBERT: Yeah, I
think that’s actually one of the things I really
like to teach my students, is that like your
instincts about things not feeling right or things
not– like you should never let the fact that you
don’t totally understand the thing get in the way of
actually asking good questions, and digging into the material. Like the idea, I do think
that all this time, people– we hear the equivalent of like
well a computer produced it, so it must be right. And I think that
scares me a lot. And I hope that one of
the reasons I really want people to learn
about technology is so that they learn that
that statement is wrong, and so that they can
unpack this idea of just because we call it
a risk score, and we use it to evaluate whether
somebody should get probation or not, and it’s no longer– and it’s a risk score instead
of sentencing guidelines. The fact that a
computer is involved doesn’t make it more accurate
or less susceptible to bias. And I think that
actually for folks who haven’t encountered
technology a lot, that can be are really revolutionary idea. Because we are meant to think– many people do
think of technology as not really being
constructed by people, and sort of not having inherent
assumptions, or biases, and being immune to
those kind of things. And so that’s why I think
technological literacy is important. But falling so in love
with the technology that you forget how real
it is for real people, and that’s what I love
about Jack’s examples, is really scary to me. And that’s one of
the things I try to impress upon my
students, with varying degrees of success. RAJ GOYLE: I think that’s such
an awesome point, because we see that bias, where
our clients will say, well, what does
your algorithm say. And then they’ll
want like the Oracle. You know, and I
was like, well, it says the Cowboys
should win this week. And then we have to
do this, like funny, because you know we’re a data
folks, we have data scientists, and then we have to play lawyer. And we have to say stop
thinking that the magic box will tell you the exact price of what
that M&A matter should cost. Think directionally. So what the data does
is give you nodes. And it tells you, you
shouldn’t overpay. That shouldn’t be a
$1.7 million matter. But it should be, based
on some nodes, 350 to 500. But there’s a reason
it could get up to 800. But you have to think a
little bit more critically. And I think the best example
of that is the FICA scores. People think there’s
some great magic Oz. And think what
the role of credit does in our society,
how hurtful it is. And that was some
loose technology that a private corporation
set up and now runs the world. And now we have to
regulate against it, we have to fight
about it, and so on, because it was a false god
that the data was pure. So I think that point
is super well taken. I see it every day
in our business. KENDRA ALBERT: Big
data is made of people. I’m full of aphorisms. RAJ GOYLE: You need to
write up a book on this. CHRIS BAVITZ: I like that. MITCH RESNICK: That’s good,
big data is full of people.

Leave a Reply

Your email address will not be published. Required fields are marked *