Patently Strategic - Patent Strategy for Startups
Patently Strategic - Patent Strategy for Startups
Software Patents: Protecting Your Digital Innovations
We're officially kicking off our second season and in this month's episode, our experts tackle software patents.
As Silicon Valley engineer and investor Marc Andreessen famously said, “Software is eating the world”. It is central to so much of today’s innovation and the growth potential is off the charts. Five companies, all with software at their core, are worth a quarter of the S&P’s entire market cap. The AI powered market will grow to $180 billion by 2025. By the end of this year, the Global Edge Computing market will reach $6.72 billion and the global AR and VR market will exceed $209 billion. By 2030, there will be 50 billion IoT devices in use worldwide. That’s 6 devices each for every man, woman and child on this planet – and all of them are powered by software.
The stakes are high, the value is soaring, and the importance of quality software patenting has never been more strategically vital. In this month’s episode, Dr. Ashley Sloat, President and Director of Patent Strategy here at Aurora leads a discussion along with our all star patent panel, digging into:
- The social and legal issues surrounding software patents
- Strategies for how to protect software innovations
- A great overview of IP implementation concerns and strategies as they pertain to outsourcing, code ownership, agile methodologies, and open source use
- An inside look at how Google manages its IP
The group also concludes with an insightful brainstorming session around two problems that plague this space. The first is around a new patent type that would help overcome many of the current prosecution and court problems faced by software patents. The second, pulling from adjacent IP law, is a mechanism for small inventors to receive compensation for using their patents, similar to how music copyright royalties are handled.
Ashley is joined today by our always exceptional group of IP experts including:
- David Cohen, Principal at Cohen Sciences
- Shelley Couturier, Patent Strategist and Search Specialist at Aurora
- Amy Fiene, Patent attorney at Vancott and adjunct professor at BYU
- Kristen Hansen, Patent Strategist at Aurora
- David Jackrel, President of Jackrel Consulting
- Daniel Wright, Partnership Manager and Patent Strategist at Aurora
** Resources **
* Show Notes: https://www.aurorapatents.com/blog/new-podcast-software-patents
** Follow Aurora Consulting **
* Home: https://www.aurorapatents.com/
* Twitter: https://twitter.com/AuroraPatents
* LinkedIn: https://www.linkedin.com/company/aurora-cg/
* Facebook: https://www.facebook.com/aurorapatents/
* Instagram: https://www.instagram.com/aurorapatents/
And as always, thanks for listening!
WEBVTT
00:05.090 --> 00:08.314
Good day and welcome to the Patenting Strategic Podcast, where we discuss all things
00:08.352 --> 00:11.230
at the intersection of business, technology and patents.
00:11.550 --> 00:15.278
This podcast is a monthly discussion amongst experts in the field of patenting.
00:15.374 --> 00:18.410
This for inventors, founders and IP professionals alike,
00:18.470 --> 00:21.958
established or aspiring. And let me tell you, it is so good
00:21.984 --> 00:25.514
to be back and officially kicking off our second season after taking a brief
00:25.562 --> 00:28.882
post holiday break. This is going to be an incredible follow up
00:28.896 --> 00:32.254
to our inaugural season. So many great topics in store for you,
00:32.292 --> 00:36.098
including upcoming deep dives into the patent Metaverse, compelling interviews
00:36.134 --> 00:39.518
with impactful up and coming inventors, and a tour of the costly
00:39.554 --> 00:42.598
misconceptions around patent ownership. But first,
00:42.684 --> 00:46.358
today our experts tackle software protection. As Silicon
00:46.394 --> 00:49.606
Valley engineer and investor Marc Andreessen famously said,
00:49.728 --> 00:53.282
Software is eating the world. It's central to so much of today's
00:53.306 --> 00:56.510
innovation, and the growth potential is off the charts,
00:56.630 --> 00:59.806
despite recent rate hikes rattling the market a bit.
00:59.928 --> 01:03.250
Microsoft, Apple, Google, Amazon, and Tesla these
01:03.300 --> 01:07.282
five companies, all with software at their core, are worth a quarter of
01:07.296 --> 01:11.450
the SP's entire market cap. It's possible that meta platforms,
01:11.510 --> 01:15.070
formerly known as Facebook, could soon be the 6th such
01:15.120 --> 01:18.958
company worth over $1 trillion. In our microcosm of Focus,
01:19.044 --> 01:22.202
these eye popping stats are no exception. The AIpowered
01:22.226 --> 01:26.038
market will grow to $180,000,000,000 by 2025. By the
01:26.064 --> 01:29.743
end of this year, the global edge computing market will reach 672 billion dollars and
01:29.760 --> 01:33.206
the global AR and VR market will exceed
01:33.278 --> 01:37.046
209,000,000,000. By 2030, there will be 50 billion IoT
01:37.118 --> 01:40.858
devices in use worldwide. That's six devices each for
01:40.884 --> 01:44.702
every man, woman and child on this planet, and all of them are powered
01:44.726 --> 01:48.494
by software. The stakes are high, the value is soaring, and the importance
01:48.542 --> 01:52.198
of quality patenting has never been more strategically vital. In this
01:52.224 --> 01:55.718
month's episode, Dr. Ashley Slote, President and director of Patent Strategy
01:55.754 --> 01:59.330
here at Aurora, leads a discussion along with our Allstar patent panel,
01:59.390 --> 02:02.786
digging into the social and legal issues surrounding software patents,
02:02.918 --> 02:06.350
strategies for how to protect software innovations, and a great overview
02:06.410 --> 02:09.718
of IP implementation concerns and strategies as they pertain to
02:09.744 --> 02:13.738
outsourcing code ownership, agile methodologies, and open source use.
02:13.824 --> 02:17.098
The group also concludes with an insightful brainstorming session around two
02:17.124 --> 02:20.590
problems that plague this space. The first is around a new patent type
02:20.640 --> 02:24.242
that would help overcome many of the current prosecution and court problems faced
02:24.266 --> 02:27.814
by software patents. The second, pulling from adjacent IP law,
02:27.912 --> 02:31.982
is a mechanism for small inventors to receive compensation for using their patents,
02:32.126 --> 02:35.818
similar to how music Copyright royalties are handled. Ashley is joined today by
02:35.844 --> 02:39.470
our always exceptional group of IP experts including David Cohen,
02:39.530 --> 02:42.998
principal at Cohen Sciences, Shelley Katarrier, patent strategist
02:43.034 --> 02:46.522
and search specialist at Aurora, Amy Fine, patent attorney at
02:46.536 --> 02:50.858
Vancouver and adjunct professor at BYU, Kristen Hansen, patent strategist
02:50.954 --> 02:54.026
and newest member of the Aurora family, David Jackroll,
02:54.098 --> 02:57.958
President of Jacquero Consulting, and Daniel Wright, partnership manager and
02:57.984 --> 03:01.930
patent strategist here at Aurora. All right, take it away, Ashley.
03:02.250 --> 03:05.662
The things I want to talk about with software is very briefly touched on the
03:05.676 --> 03:08.400
social legal issues because you all know what those are,
03:09.030 --> 03:12.298
but just more to kind of bring everybody with the second part
03:12.324 --> 03:16.078
of the presentation and then how to protect software innovations. I think we're all
03:16.104 --> 03:19.342
pretty clear on what those options are. But again, just a brief touch on that.
03:19.356 --> 03:23.450
But then the bigger part that I want to talk about is implementation
03:23.510 --> 03:28.570
concerns of strategies. Some of these things were things that I work,
03:28.740 --> 03:33.002
I think cursory thoughts, but digging
03:33.026 --> 03:36.298
into them where you can kind of see some of the nitty gritty of it.
03:36.324 --> 03:39.706
And so I think that's this last piece is kind of some of
03:39.708 --> 03:42.958
the more eye opening or interesting topics for
03:42.984 --> 03:46.634
conversation. So, of course, with software patents,
03:46.682 --> 03:50.280
I think some of the interesting dynamics are that
03:50.730 --> 03:54.502
socially they're kind of sorted to camp some that fully believe
03:54.576 --> 03:57.874
that software should be patent protected. And there's another group of
03:57.912 --> 04:01.354
individuals who think that software should be an open
04:01.392 --> 04:05.170
source. Basically, both want to promote access to software,
04:05.490 --> 04:09.010
but they basically believe that it should be done about two different ways. One is
04:09.120 --> 04:12.082
make it open source and anybody can build upon it.
04:12.276 --> 04:16.058
Easy peasy. The other one believe that you file
04:16.094 --> 04:19.270
patents and then you can still give access to people,
04:19.320 --> 04:22.210
but you still are protecting your base innovation.
04:23.310 --> 04:26.722
These groups want kind of the same thing, but they also
04:26.796 --> 04:30.082
want to go about it in two different ways. And then legally, obviously,
04:30.156 --> 04:33.694
there's some issues around the Alice decision and that
04:33.732 --> 04:37.654
some of the complexities and uncertainty that that put into
04:37.692 --> 04:41.338
the patent system around software patenting. And then of course, all the case law from
04:41.424 --> 04:44.678
the court system, the Federal Circuit has only increased
04:44.714 --> 04:48.518
the ambiguity and then either T tab obviously adds additional ambiguity
04:48.554 --> 04:52.294
as well because handling things in a little bit different way, a little bit different
04:52.332 --> 04:54.982
perspective, and trying to make heads or tails of it. And really at the end
04:54.996 --> 04:58.970
of the day, we do have some broader overarching ideas.
04:59.030 --> 05:02.542
But still there are some cases out there where even some
05:02.556 --> 05:06.682
of the ones that I read, I look at some of the ones that are
05:06.816 --> 05:09.958
invalidated. I'm like, that's ridiculous. And then you look at some of the other
05:09.984 --> 05:12.746
ones that were held up and you're like, that's ridiculous.
05:12.938 --> 05:17.722
So there's still some weird stuff where I
05:17.736 --> 05:20.938
don't feel there's a good consensus between the courts about how that should all be
05:20.964 --> 05:24.778
handled anyways. Of course, we all do know is that
05:24.924 --> 05:28.042
there's a lot of ineligibility happening in the software space.
05:28.116 --> 05:32.182
Lots of times are being found Invalid either in
05:32.256 --> 05:35.558
the Uscto or at the Pizza or Federal Circuit based on all levels.
05:35.594 --> 05:39.074
There's some degree of, I would say, higher ineligibility
05:39.122 --> 05:42.418
than eligibility with software pens. I think it does beg the
05:42.444 --> 05:45.922
question of how to remedy that. I do have some thoughts, like I
05:45.936 --> 05:49.078
said at the end, to talk through. But of course. And then,
05:49.104 --> 05:52.970
as we all know, the main problem is that software
05:53.030 --> 05:56.330
does fall into one of the four categories.
05:56.450 --> 06:00.110
Process, machine, manufacturer comes to matter, obviously, likely a machine
06:00.170 --> 06:03.638
or a process. The courts have this implicit exception,
06:03.794 --> 06:07.440
one of them being abstract ideas. And software tends to fall into that.
06:08.310 --> 06:11.470
But despite all that, even with that
06:11.520 --> 06:14.160
history, even with the social pressures that are happening,
06:14.730 --> 06:18.422
software is routinely protected through utility patents. It's routinely protected
06:18.446 --> 06:21.974
through design patents like the ornamental
06:22.022 --> 06:25.202
aspects of animations or graphical user interfaces.
06:25.286 --> 06:29.822
You can do trademarks for your software. You can even do trade secret depending
06:29.966 --> 06:33.262
and then Copyright for unique software code expressions. So I
06:33.276 --> 06:36.898
think it's still very protectable. But I think where
06:36.924 --> 06:40.498
I want to go with this presentation now is some of
06:40.524 --> 06:44.222
the intricacies of copywriting software versus
06:44.306 --> 06:47.422
patenting software. And what are some of the gotchas? Because these are some
06:47.436 --> 06:49.500
things like I said as I dug into them more,
06:51.090 --> 06:54.502
there were some things where I think if a person or
06:54.576 --> 06:58.838
inventor or company is unwary, they could get caught in a problematic
06:58.874 --> 06:59.460
situation.
07:01.810 --> 07:05.260
Ashley, can I interrupt you real quick?
07:05.770 --> 07:09.290
You listed Gui's under design patents, and you can do that.
07:09.340 --> 07:12.854
But graphical user interfaces can also be under utility because
07:12.892 --> 07:16.370
there's so much utility underneath them, the algorithms and stuff.
07:16.420 --> 07:19.958
So it could be either. Yeah, absolutely. That's a great point.
07:20.104 --> 07:23.822
Absolutely. And I think that's where the interesting line
07:23.896 --> 07:27.254
blurs a little bit between utility and design,
07:27.352 --> 07:30.726
because separating the ornamental aspect
07:30.798 --> 07:34.794
from the functional aspect too. And how do you tout the functional aspect
07:34.842 --> 07:39.270
versus the ornamental aspect, any kind of consumer facing documentation
07:39.330 --> 07:42.050
and things like that, and making sure you're kind of following,
07:42.970 --> 07:45.460
being careful about how you do that.
07:46.930 --> 07:50.282
So obviously, you all probably know this, but copyrights are meant for
07:50.296 --> 07:53.654
the literal expression of the
07:53.692 --> 07:57.114
idea. So like, the little expression of software, code is more protected
07:57.162 --> 08:01.362
by Copyright, and so it's guarding against exact replication or derivations
08:01.446 --> 08:04.514
of that and protect the broader idea. Right.
08:04.552 --> 08:08.546
That no matter how it's implemented, if it's different languages or maybe
08:08.608 --> 08:11.906
different ways of reducing it to practice, it doesn't really
08:11.968 --> 08:15.758
matter unless your claims have that specific reduction to
08:15.784 --> 08:18.890
practice. Right. Unless you very specifically say you have to do X, Y, and Z.
08:18.940 --> 08:21.280
And there's only one way to do X, Y, and Z with code.
08:22.510 --> 08:26.342
So some of the interesting thing is that I think agile development has
08:26.416 --> 08:30.206
brought in a lot of interesting opportunities for the patent process.
08:30.388 --> 08:33.486
I don't know how many of you are familiar with the agile
08:33.678 --> 08:37.686
process, but it obviously encourages fast paced work cycles.
08:37.698 --> 08:41.238
So historically they had like, I think what they called the waterfall development,
08:41.274 --> 08:45.198
where you basically the
08:45.224 --> 08:48.834
team that was deciding what the product features would be. They would basically plan
08:48.872 --> 08:52.218
out the entire product and then throw it over the wall to
08:52.244 --> 08:55.878
the developers and then, you know, some years later the developers would emerge with a
08:55.904 --> 09:00.020
product and that product didn't necessarily account for
09:00.770 --> 09:04.642
real time feedback or usability or consumer feedback
09:04.726 --> 09:08.542
or changes in the product pathway or anything like that. So it kind of caused
09:08.566 --> 09:11.826
this problem that you needed to iterate on more
09:11.888 --> 09:15.258
smaller segments of software to get it to be where you want it to be
09:15.284 --> 09:18.622
for consumers. And so it's faster pace, work cycles,
09:18.706 --> 09:22.746
rapidly evolving objectives. You can change what each product,
09:22.868 --> 09:26.274
each segment of the product, how it behaves and how it looks based
09:26.312 --> 09:29.554
on consumer feedback to the prior iteration or additional
09:29.602 --> 09:33.450
feedback that happens later on. So it's just easier to kind of evolve the product
09:33.500 --> 09:37.654
over time. They also do very stringent resource prioritization.
09:37.762 --> 09:42.510
So putting developers on that
09:42.560 --> 09:46.602
cycle at that time and then reallocating them to the next one and tying it
09:46.616 --> 09:50.434
together and then incremental completion.
09:50.542 --> 09:54.080
So again, continuous delivery of smaller portions. And then
09:55.130 --> 09:58.818
I think this creates a problem in the sense that
09:58.964 --> 10:02.274
it's harder to know when there's enough there to protect you're only seeing
10:02.312 --> 10:06.258
these little bits happening from a software development cycle from
10:06.284 --> 10:07.710
an IP perspective,
10:10.010 --> 10:13.278
sorry, without the forest, it might be hard to see the trees. And so being
10:13.304 --> 10:17.134
able to kind of see both aspects of it through this incremental development
10:17.242 --> 10:20.914
and making sure that you're abreast with the development team as they're developing, as objective
10:20.962 --> 10:24.778
change, as prioritization change, as features change, making sure that you're
10:24.814 --> 10:28.160
kind of on board with the development team to see what happens when.
10:29.210 --> 10:32.466
And then, of course, there's the risk with early disclosure and customer
10:32.528 --> 10:36.454
feedback, with the shorter experience, you're getting more feedback potentially
10:36.502 --> 10:40.518
at shorter intervals. You're pushing it out to consumers potentially. In theory, this is all
10:40.544 --> 10:44.086
experimental use, but in the event that it's not that you're
10:44.098 --> 10:47.890
actually selling it and still doing customer testing and feedback,
10:48.010 --> 10:51.958
this can also present problems for clients as well. And then lastly,
10:51.994 --> 10:55.146
of course, we always have I think we all experience this a
10:55.148 --> 10:58.902
lot. Inventors always want to self censor and say, well, it's just a
10:58.916 --> 11:02.662
simple thing like everybody does this in software. Everybody uses
11:02.806 --> 11:06.678
this code to do this, everybody does that. But again, in the bigger pie,
11:06.764 --> 11:10.482
in the bigger forest, does that the way they've used that code make a
11:10.496 --> 11:13.220
difference? Does their implementation make a difference?
11:14.870 --> 11:18.262
So here's kind of one example of agile software.
11:18.346 --> 11:22.134
And so obviously many companies do this differently. But what I think
11:22.172 --> 11:25.582
is probably the most useful places
11:25.606 --> 11:28.674
for IP input is that you're going to look for it. Do you have the
11:28.712 --> 11:32.514
requirements and user input? Of course, us being involved in that
11:32.552 --> 11:35.934
process, I think is really important because you're starting to understand what's the problem
11:35.972 --> 11:39.114
that they're solving, what's the product that they want to build, what do the
11:39.152 --> 11:42.718
users want out of this product? But then it's really the planning
11:42.754 --> 11:45.610
phase and they're starting to think about how they're going to reduce it to practice,
11:45.670 --> 11:49.446
or what are the chunks, the different sprints that they're going to do?
11:49.568 --> 11:53.118
I think it's a great point for us as practitioners to step in and
11:53.144 --> 11:57.030
start to identify whether might be protectable elements. And then of course,
11:57.080 --> 12:00.426
as development sprints happen, you're going to start to see actually how
12:00.488 --> 12:04.266
things play out in terms of how they're actually developed. So you may say that
12:04.448 --> 12:07.746
the problem to solve is this or we want to solve it via that.
12:07.808 --> 12:11.598
But of course, as all software development can
12:11.624 --> 12:15.126
be really difficult at times, they're going to encounter different problems,
12:15.188 --> 12:18.942
different technical problems that we can evaluate from
12:18.956 --> 12:23.302
a patentability perspective at each development for under each development cycle.
12:23.446 --> 12:26.322
And then lastly with user review, if they're pushing it out, to use it for
12:26.336 --> 12:30.142
beta testing, if they're having internal people do it, or however they're
12:30.166 --> 12:32.878
getting that test group feedback,
12:33.034 --> 12:35.962
obviously you're going to want to do that before they push it out and you're
12:36.046 --> 12:39.198
definitely going to want to do it before they release it. And then
12:39.224 --> 12:43.110
lastly, I think the retrospect is interesting because that
12:43.160 --> 12:46.554
allows everybody to weigh in on how things went.
12:46.592 --> 12:49.686
Do they go well, do they not go well, do things behave like they wanted
12:49.748 --> 12:53.190
to or what were alternatives to what were done?
12:53.360 --> 12:56.826
And so I think even in the retrospective phase could be a
12:56.828 --> 13:00.222
valuable place to see if there is it just because I
13:00.236 --> 13:03.574
think that could be an opportunity to define what alternatives
13:03.622 --> 13:07.122
there may be. I noticed different developers are
13:07.136 --> 13:11.734
going to tackle a problem in a different way. And so through that retrospective,
13:11.842 --> 13:15.498
different developers hopefully speak up and say, I like the way you did
13:15.524 --> 13:18.810
that, but I would have done it this way differently.
13:20.150 --> 13:23.622
So that's just one example where I think we can kind of play into the
13:23.636 --> 13:24.930
software development cycle.
13:26.450 --> 13:29.886
And so then again, like just going after this
13:30.008 --> 13:33.498
further is that I think from a software, if you look
13:33.524 --> 13:37.606
at each one of these data points as a sprint, you have four sprints shown
13:37.738 --> 13:40.878
you really can have multiple provisions being filed and then
13:40.904 --> 13:43.890
all being wrapped up into one utility right before release.
13:44.270 --> 13:48.140
And also you could of course abandon different Provisionals at different times
13:48.530 --> 13:53.742
without really any penalty. So I think being
13:53.756 --> 13:57.198
able to do the smaller chunks of software patents, I think with the
13:57.224 --> 14:00.582
bigger goal in mind, with a better, bigger product in mind, with the bigger problems
14:00.656 --> 14:04.110
to be solved in mind with really making
14:04.160 --> 14:07.038
sure you're seeing the forest. Yeah,
14:07.124 --> 14:10.354
I agree, Ashley. This is a good strategy,
14:10.522 --> 14:14.526
especially with the Provisionals earlier, because the biggest buzz saw that I see
14:14.588 --> 14:18.862
clients run into with agile development in house is losing
14:18.946 --> 14:21.870
foreign protection because you've inadvertently disclosed.
14:22.490 --> 14:25.662
So you get to the end or you get
14:25.676 --> 14:29.614
to the middle and you can't actually go after foreign rights because you've inadvertently
14:29.662 --> 14:33.214
disclosed or behind a flag or whatever. And in software
14:33.262 --> 14:36.560
it's really difficult to prove it was experimental use.
14:36.890 --> 14:40.206
It's so easy to get to. It's so easy to release. Right?
14:40.388 --> 14:44.382
Right. Do they do a lot of stuff
14:44.396 --> 14:47.840
in their user agreements that say anything like that?
14:51.090 --> 14:54.540
Yeah, as long as you're under NDA and I believe,
14:56.010 --> 15:00.638
of course, no sales. But as long as you're under NDA and you haven't sold
15:00.674 --> 15:01.260
it,
15:05.250 --> 15:08.678
I'm trying to think what else, but definitely NDA,
15:08.714 --> 15:12.254
you want to be careful because that isn't truly a release
15:12.302 --> 15:16.042
to public. Right. So then you can protect a
15:16.056 --> 15:19.570
lot more further back. But if you have these Provisionals going along,
15:19.740 --> 15:22.714
as long as you don't go past your one year of that prior one,
15:22.752 --> 15:26.174
you can go back and get all of that protection in a foreign
15:26.282 --> 15:29.630
space. The bigger
15:29.690 --> 15:33.118
problem, though, is they often want to release full product
15:33.264 --> 15:36.742
at each little tick because it's usually an app or
15:36.756 --> 15:40.054
it's something that does five things
15:40.092 --> 15:43.246
and they're working on one of five. Right. So that's kind of hard
15:43.308 --> 15:46.378
even to write a provisional about a simple thing if it doesn't seem all that
15:46.404 --> 15:50.398
inventive at first glance. Right. So I
15:50.424 --> 15:54.060
don't know. I mean, maybe it's a cheaper than normal provisional. I don't know.
15:54.630 --> 15:57.418
Do you see a lot of in that world? Do you see a lot of
15:57.444 --> 16:00.982
continuations in part before publication dates and stuff like
16:00.996 --> 16:03.480
that to try to capture? I do.
16:03.930 --> 16:07.378
Well, you can do it that way. And it's kind of unfortunate because you do
16:07.404 --> 16:10.606
lose some stuff. But more often than not,
16:10.668 --> 16:13.798
I see, because they've taken an agile approach to
16:13.824 --> 16:17.218
software, they've lost a huge chunk of their
16:17.244 --> 16:20.842
invention. And so they have to take this point and go forward and
16:20.856 --> 16:24.394
they can only invent and claim forward. Right.
16:24.552 --> 16:27.818
So sometimes that means you only get a couple of methods.
16:27.854 --> 16:31.090
You don't get a fully functional piece of software,
16:31.530 --> 16:33.900
you just get a few algorithms. Right.
16:36.490 --> 16:39.998
It's tougher. But if your people know, if you tell
16:40.024 --> 16:43.358
them, you're going to run into form filing problems, if you don't disclose early and
16:43.384 --> 16:47.066
often to your IP team, if you don't think about these things
16:47.128 --> 16:50.342
early, it could be a problem. And basically you can look at
16:50.356 --> 16:53.766
this from a PowerPoint at a product developer
16:53.838 --> 16:57.400
product management level. When those steps are being
16:57.970 --> 17:02.118
presented to the company internally, an IP attorney
17:02.154 --> 17:05.690
or IP agent can take that slide deck and pull out what they think
17:05.740 --> 17:09.450
might be inventive and then have those quick inventor discussions
17:09.570 --> 17:12.758
to see if they are indeed something new that is not
17:12.784 --> 17:16.334
yet released. Right. That's what we did at
17:16.432 --> 17:19.862
Scientific Atlanta. We had every product
17:19.936 --> 17:24.338
management meeting. We had the It guy right
17:24.364 --> 17:27.974
there. He was part of the meeting. So we would
17:28.012 --> 17:31.698
even as a product manager, that was our goal,
17:31.794 --> 17:35.630
to look at anything inventive and then send it off immediately.
17:37.570 --> 17:41.282
Okay. So then inside Council would send that to outside
17:41.356 --> 17:44.586
counsel, and outside Council can be paid to do the due
17:44.598 --> 17:48.530
diligence to follow up to do some interviews, and sometimes it doesn't turn into
17:48.580 --> 17:52.638
an application, but it's still a couple of hours of analysis and interviews.
17:52.734 --> 17:55.480
It's still a feed structure of work,
17:56.350 --> 17:59.766
even if it doesn't turn into a patent.
17:59.838 --> 18:04.120
But I think the hard part is as giving
18:04.870 --> 18:08.402
companies and inventors to be
18:08.416 --> 18:12.042
more inclusive in the sense of bringing
18:12.186 --> 18:15.950
legal counsel on for the ride a little bit. I think it's just an extra
18:16.120 --> 18:20.438
step and extra burden, but I think the payoff can be really big as well.
18:20.464 --> 18:24.918
You're both pointing out some
18:24.944 --> 18:28.938
of this. I just go to the user agreement stuff. It's having some
18:28.964 --> 18:32.718
of that NDA language in there, and that any
18:32.744 --> 18:36.318
of the feedback that your clients going to be receiving can
18:36.344 --> 18:39.078
be incorporated into the product, can be used in the product,
18:39.164 --> 18:42.642
and without royalties or
18:42.656 --> 18:46.294
payments, you are going to be getting feedback from consumers around different product sets
18:46.342 --> 18:49.590
and different improvements we made. And so we definitely
18:49.640 --> 18:52.758
want your client to own that. So I think, again, we just not like
18:52.784 --> 18:59.074
we're going to drop this language for them as agents. But I think making
18:59.112 --> 19:02.782
sure your client is thinking through all of the contracts. I think the more and
19:02.796 --> 19:05.710
more I deal with small companies or solos,
19:06.210 --> 19:10.306
the more I'm surprised that a
19:10.308 --> 19:12.886
lot of them don't have basic agreements. I spoke to a gentleman the other day
19:12.948 --> 19:16.510
who has been working with another individual. This is a software. This is
19:16.560 --> 19:20.558
an actual device, but hadn't
19:20.594 --> 19:23.734
done an agreement with that gentleman with his company. And so if that person
19:23.772 --> 19:28.270
became disgruntled later and decided that whatever
19:28.320 --> 19:31.462
he brought to the table was valuable enough and he didn't want to assign it
19:31.476 --> 19:35.770
or something, now you have this weird issue. And so I think really
19:35.820 --> 19:38.618
reinforcing the clients the value in contracts,
19:38.654 --> 19:41.722
NBA contracts, work to hire all that kind of stuff and making sure
19:41.736 --> 19:44.566
they have it's not like we have to do that. We're not going to be
19:44.568 --> 19:48.240
doing that work. But I think we all view ourselves as
19:49.050 --> 19:52.814
responsible for them, hopefully conducting their business in a responsible
19:52.862 --> 19:56.458
way. And sometimes that means reminding them
19:56.484 --> 19:59.340
that they need certain contracts and stuff like that.
19:59.850 --> 20:01.920
Sure. They haven't gone yet.
20:04.170 --> 20:07.882
Okay, some open source considerations. This is probably the most, like eye opening thing to
20:07.896 --> 20:11.554
me, I guess a little bit. I guess not like the copy left stuff
20:11.592 --> 20:13.800
and things like that. But anyway,
20:15.750 --> 20:19.514
obviously there's open source licenses. These are royalty free Copyright
20:19.562 --> 20:23.198
licenses, branching third parties the rights to copy, modify, and distribute
20:23.234 --> 20:26.722
computer code with strings attached that are dependent on the type of
20:26.736 --> 20:30.278
source code that you're using. Most of us are probably familiar with copy
20:30.314 --> 20:34.934
left, where essentially if you incorporate copyrighted
20:34.982 --> 20:38.398
software into your software, then you have
20:38.424 --> 20:41.818
to distribute that software with the same parameters as the
20:41.844 --> 20:45.298
copy last source code. So essentially you have to make
20:45.324 --> 20:48.538
your software open source, which I'm sure your client would juggling that
20:48.564 --> 20:51.922
like or maybe they do have certain portions of it, but you have to make
20:51.936 --> 20:54.970
your whole base open source, and that'd be ideal.
20:56.370 --> 20:59.890
And I think other questions that we get from clients is
20:59.940 --> 21:03.062
that whether open source can be patented.
21:03.206 --> 21:06.374
And so it can be and is often patented.
21:06.542 --> 21:10.330
And so it just depends on how you're using it,
21:10.380 --> 21:15.110
how you're extending beyond that open source, because obviously that open source software
21:15.170 --> 21:18.802
code is available to the public. So I
21:18.816 --> 21:22.142
mean, I view it as a published prior art document for all intents
21:22.166 --> 21:25.390
and purposes. And so how is your client actually using
21:25.440 --> 21:28.970
that open source code? How are they extending beyond it? How are they incorporating
21:29.030 --> 21:32.422
into their product? How are they using it, all that kind
21:32.436 --> 21:36.574
of stuff. But there are built
21:36.612 --> 21:39.502
into open source. Again, it depends on the license. So you want your clients to
21:39.516 --> 21:42.960
be able to read that and weigh the
21:44.790 --> 21:48.758
implications of that license. There's a lot of times some kind of express or implied
21:48.794 --> 21:52.490
license grant in open source software. So if your client is incorporating
21:52.550 --> 21:55.966
it, they're likely fine to do that. But again, they just need to understand
21:56.028 --> 21:58.750
what the licenses and the requirements of that license.
22:00.150 --> 22:03.730
Interestingly, there's also patent retaliation clauses.
22:04.710 --> 22:08.834
So if your client attempts to assert
22:08.942 --> 22:12.422
infringement against the OSS developer, there's causes
22:12.446 --> 22:15.322
in there that can prevent that kind of activity. So again, it's kind of a
22:15.336 --> 22:18.358
two way street. Yes, you can use this. I'm also protected from you
22:18.384 --> 22:20.460
doing something selling later against me.
22:23.050 --> 22:26.654
So the way I view it, for whatever reason,
22:26.692 --> 22:30.074
this last year, I've seen a lot of this come through. And that's also probably
22:30.112 --> 22:33.880
why we prepared this talk for another group is just that,
22:34.750 --> 22:39.002
again, from a contract perspective, you see a lot of clients having
22:39.136 --> 22:43.022
software developed like non technical founders or founders who
22:43.036 --> 22:46.878
don't have the time to actually do the technical aspects of their product. They outsource
22:46.914 --> 22:50.318
it, but then they don't do the contracts needed to
22:50.344 --> 22:53.882
protect themselves. Because I'm sure you all appreciate when
22:53.896 --> 22:57.674
you come into relationship with a Dev shop, they have
22:57.712 --> 23:00.734
their own code, right? They have their code bases that they use because
23:00.772 --> 23:03.350
a lot of times they build kind of similar products for a lot of different
23:03.400 --> 23:07.058
businesses. If they're all doing mobile applications, they may have like a base
23:07.204 --> 23:10.958
mobile application data code set that they use, and then they build upon that
23:10.984 --> 23:14.378
for specific clients. There's also open source that they need
23:14.404 --> 23:17.438
to be pulling from. And I don't know any developer that doesn't pull from open
23:17.464 --> 23:20.786
source because it'd be silly not to because it just gives you such
23:20.848 --> 23:24.722
a bump in how much code you can develop and how quickly you can
23:24.916 --> 23:28.574
then also as a part of the relationship that you're going to
23:28.612 --> 23:32.066
form with a Dev shop, there's also going to be new code generated that specific
23:32.128 --> 23:35.750
to your product. And so the product obviously lies somewhere in
23:35.860 --> 23:39.134
that Venn diagram of the Dev shock code, the source code and
23:39.172 --> 23:42.710
your new code. And so what I think people
23:42.760 --> 23:46.442
don't fully always realize is that you
23:46.456 --> 23:49.910
need to have contracts in place to make sure that
23:49.960 --> 23:52.974
the new code is controllable by your clients,
23:53.142 --> 23:56.378
making sure that whatever contract is created, make sure that any new code
23:56.404 --> 23:59.730
that's developed as part of that product is controllable by your client
23:59.790 --> 24:03.854
or has the It rights are assigned to your client. You have to make sure
24:03.892 --> 24:07.914
that they're complying with the open source licensing
24:07.962 --> 24:11.486
structure. And then there's lots of the Dev shop. The desktop has
24:11.488 --> 24:14.750
to be brought along in the process of making sure that they're complicit in whatever
24:14.800 --> 24:19.022
agreement is set up. We had a client come to us where they
24:19.036 --> 24:22.418
had come to us right before a conversion event, and they
24:22.444 --> 24:26.800
had a product developed. They paid the development company
24:27.310 --> 24:30.998
for the software development, but this company had
24:31.084 --> 24:34.478
obviously based code that they were developing off of. They do a lot
24:34.504 --> 24:37.898
of like ML stuff for a lot of different companies and there was
24:37.924 --> 24:41.538
no development agreement up front. And so then when it came time to conversion,
24:41.574 --> 24:45.254
when we're trying to figure out assignments and all that kind of stuff, the Dev
24:45.292 --> 24:48.602
shop says, no, it's built on our
24:48.616 --> 24:51.890
base code, we own it. Our clients are saying, no,
24:52.060 --> 24:55.600
I pay you for all of this, we own it.
24:55.990 --> 24:59.282
And it was an amicable relationship where I think even the desk shop had
24:59.356 --> 25:03.086
invested in this company and they had a really hard time
25:03.148 --> 25:07.074
coming to terms. And ultimately the terms became too, which was not ideal,
25:07.122 --> 25:10.638
but better than nothing was that the Dev shop and the client
25:10.674 --> 25:14.438
company became co owners of the IP. But from a
25:14.464 --> 25:18.270
licensing investor acquisition perspective,
25:18.330 --> 25:21.794
it could be really problematic that that structure exists. And so
25:21.832 --> 25:25.274
it just really underscore the need for there
25:25.312 --> 25:28.514
to be some kind of agreement upfront that says, we already understand Dev shop,
25:28.552 --> 25:31.946
but you're bringing this to the table. You understand that we're bringing this to the
25:31.948 --> 25:36.074
table and paying you for certain things. And then the thing that's developed out
25:36.112 --> 25:39.554
of this relationship, who owns that, what does that mean?
25:39.652 --> 25:43.602
Whose name is listed on IP? How is that controllable?
25:43.746 --> 25:47.066
Whatever. And you'd be surprised. At least I was.
25:47.128 --> 25:49.658
I guess this last year has been a really big eye opener for me about
25:49.684 --> 25:53.078
how many companies don't set
25:53.104 --> 25:56.766
up this way. They don't do these diligence
25:56.898 --> 26:00.242
stats early on. They create problems that we have to fix later, which can be
26:00.256 --> 26:01.000
really difficult,
26:05.690 --> 26:07.760
actually, before you go on,
26:08.630 --> 26:12.426
which is reminding me of the recent case with
26:12.488 --> 26:16.354
Java and Google and Android.
26:16.522 --> 26:20.094
Yeah, I know that decision was really
26:20.132 --> 26:24.174
controversial, but it
26:24.212 --> 26:28.078
seems very tricky. The previous Venn diagram
26:28.114 --> 26:32.238
you had, especially with the open source compliance stuff,
26:32.324 --> 26:36.102
it seems I don't do a ton of software work,
26:36.296 --> 26:39.790
but in that decision and some other conversations
26:39.850 --> 26:43.498
we've had and such, it seems that a lot of Gray
26:43.534 --> 26:47.734
area. Do you agree with that? Is everybody familiar
26:47.782 --> 26:50.718
with Christian? Yeah, yeah,
26:50.864 --> 26:55.638
actually, I have prosecuted for Google since about 2006.
26:55.784 --> 26:59.480
Nice. And they are very touchy about this,
26:59.930 --> 27:03.874
so much so that they have a lot of partnerships.
27:03.922 --> 27:07.230
They do a lot of this sort of combination where the Dev shop
27:07.280 --> 27:10.482
code isn't a Dev shop, it's another company or it's another set of
27:10.496 --> 27:13.818
inventors, or it's another product or piece that they're going
27:13.844 --> 27:17.370
to acquire or they're looking at acquiring and they will actually
27:17.420 --> 27:19.230
go together and file IP.
27:20.990 --> 27:25.626
They do that out of hope to get something usually
27:25.688 --> 27:28.842
from the engineering team and the development side and
27:28.856 --> 27:32.082
to get that skill set in house rather than the
27:32.096 --> 27:35.394
IP that would come from it immediately. Right.
27:35.492 --> 27:39.102
So future IP that maybe isn't quite related to
27:39.116 --> 27:42.330
that or is a little bit different then. They do have rights to that right
27:42.380 --> 27:47.060
after they've acquired. But from
27:48.830 --> 27:52.834
an open source standpoint, they just pay attention
27:53.002 --> 27:58.450
when it comes to IP. They're very
27:58.500 --> 28:02.314
well versed in what they can and cannot do. So much
28:02.352 --> 28:06.038
so that even their engineers have a clue about when they're
28:06.074 --> 28:09.602
using open source, when they're creating open source,
28:09.746 --> 28:13.402
what they can release, what can go behind a flag, for example,
28:13.476 --> 28:16.354
on Chrome and what cannot. They're huge,
28:16.512 --> 28:18.790
hugely paying attention.
28:19.650 --> 28:22.080
But even that, it happens to them.
28:24.210 --> 28:26.914
The thing is to avoid it, but they also want to be part of that
28:26.952 --> 28:30.970
industry. So it's really hard to avoid. Yes. So from your
28:31.080 --> 28:34.860
experience and perspective, I don't know if you know
28:35.490 --> 28:39.250
this Google versus Oracle decision from April,
28:40.350 --> 28:41.530
but essentially,
28:44.830 --> 28:48.640
I guess it was the Supreme Court. Yes, the Supreme Court said that
28:49.090 --> 28:52.322
Google was not infrinting or rather
28:52.396 --> 28:57.002
that Oracle didn't have a
28:57.016 --> 29:00.414
right or that the patent was valid, I guess is maybe what the decision
29:00.462 --> 29:03.582
was. Because even though it was hundreds and hundreds
29:03.606 --> 29:07.358
of lines, maybe thousands of lines of code, it was somehow like
29:07.504 --> 29:11.642
not integral or like because it was the
29:11.836 --> 29:15.654
right. Sorry. Because they said it was the APIs,
29:15.762 --> 29:19.062
which are like not they never spoke
29:19.086 --> 29:23.054
to the copyrightable of APIs. They basically said because it
29:23.092 --> 29:25.130
was like an organizational scheme,
29:26.830 --> 29:30.674
even though, like I said, it was like a huge percentage of the actual code
29:30.712 --> 29:34.338
base. It was insane. But yeah, it was an API
29:34.374 --> 29:37.814
and not something that they were privy to.
29:37.852 --> 29:39.880
They were just using it. Right.
29:44.270 --> 29:47.466
Oss compliance. I mean, by your
29:47.528 --> 29:50.622
understanding and everything is Google
29:50.696 --> 29:54.310
Incompliance. Were they doing all that properly and Oracle
29:54.490 --> 29:57.882
was in the wrong? I don't know,
29:58.076 --> 30:01.350
because I know how Google normally works and they normally would
30:01.400 --> 30:04.758
comply with any and all things like that because they would usually
30:04.844 --> 30:07.820
be the big dog that someone was trying to sue. Right.
30:08.390 --> 30:11.694
So they work really hard to not work
30:11.732 --> 30:14.778
in that space if they don't have to, but to also share when they
30:14.804 --> 30:15.380
can.
30:18.630 --> 30:22.738
But honestly, it's a matter of who
30:22.764 --> 30:26.074
is using what, how they're using it and how much access they have
30:26.112 --> 30:29.520
to it as a matter of write.
30:29.970 --> 30:33.806
So if it's true open source, it's usable
30:33.878 --> 30:37.570
and it's viewable. Right. I can see the code in that API,
30:38.070 --> 30:41.498
which is usually not the case, but true open source
30:41.534 --> 30:43.380
code. That's how I look at it.
30:44.910 --> 30:49.138
Interesting. Okay. But that's also if
30:49.164 --> 30:51.598
you look at kind of what went on there and what has gone on in
30:51.624 --> 30:55.618
similar situations, that's how standards are born, too. Sometimes when
30:55.644 --> 30:58.798
it becomes the standard, people are like, well, we're not going to pay for it.
30:58.824 --> 31:02.594
Right. Just the standard. And it's just a little different path.
31:02.642 --> 31:06.526
Right. Well, I think it gets a little tricky, at least from
31:06.588 --> 31:10.126
the clients I've dealt with. Is that how do you
31:10.248 --> 31:13.594
ensure because a lot of times you have a software, you got to express the
31:13.632 --> 31:18.130
base, you got to express the base technology
31:18.300 --> 31:21.658
and the actual product. Right. Like, you can't just do it at such a high
31:21.684 --> 31:25.318
level that you're like, here's the product that I've shopped made for
31:25.344 --> 31:30.262
us, and here's just that piece. A lot of times you need to like,
31:30.276 --> 31:33.442
in this case, it was like base ML programs and things like that that were
31:33.456 --> 31:36.670
being used in an ML centric
31:37.110 --> 31:40.678
applied ML case. You still have to go into the
31:40.704 --> 31:43.714
details of the ML. Right.
31:43.752 --> 31:47.902
How it's trained, how it's used, how it's whatever. And so I think a
31:47.916 --> 31:51.334
lot of the great areas speaking to your point, David, that came into this case
31:51.372 --> 31:55.706
was that in order for us to effectively claim
31:55.898 --> 31:59.398
and draft about the new code that
31:59.424 --> 32:01.150
was developed for our clients,
32:02.370 --> 32:05.974
we had to disclose to some degree the
32:06.012 --> 32:12.214
code that was the base code that the Dev shop brought and
32:12.372 --> 32:15.720
future proof for that Dev shop because the Dev shop doesn't want to. Later,
32:16.650 --> 32:20.242
you have claims that are broad that could be applied to any case when
32:20.256 --> 32:24.358
they've used that code for other clients. So you
32:24.384 --> 32:27.694
do have to enable everything. But sometimes
32:27.852 --> 32:30.982
your machine learning is not
32:31.116 --> 32:34.798
novel because it's not a new network. It's kind of
32:34.824 --> 32:38.078
off the shelf, if you will. It's not newly trained
32:38.114 --> 32:40.978
except for it happens to be on your data. Right. But it's not a new
32:41.004 --> 32:44.818
kind of data or it's not a new problem solved. But you
32:44.844 --> 32:48.778
can claim that you're receiving this content, gathering this
32:48.804 --> 32:51.982
data, you're putting it into some sort of
32:51.996 --> 32:55.342
neural network system or some sort of machine learning black box,
32:55.536 --> 32:58.858
and something is created out of that and you do something with
32:58.884 --> 33:02.640
what's created out of that. So this kind of another way to claim that
33:03.090 --> 33:06.418
without saying I've invented this neural network, I've invented this
33:06.444 --> 33:08.580
machine learning process, right?
33:09.930 --> 33:13.630
Yeah, for sure. There's only some ways. But my experience,
33:13.680 --> 33:17.580
too, with the Patent office, with machine learning and AI, though, too, it's just like
33:18.510 --> 33:22.418
a generic recitation of an offtheshelf Mlai
33:22.514 --> 33:26.282
is not enough. They still want to know the details,
33:26.366 --> 33:29.038
even if it is off the shelf. They want to know the details of how
33:29.064 --> 33:32.530
it's done either in the SAG and then maybe even to some
33:32.580 --> 33:36.398
degree in the claims. And so it gets that weird in that client
33:36.434 --> 33:39.970
case, you may be the co ownership with appropriate because at some point
33:40.020 --> 33:43.814
the client decides to be super broad and just say this ML
33:43.862 --> 33:47.698
system, then the other party can say we
33:47.724 --> 33:50.160
own that patent because that was our base pack.
33:51.450 --> 33:55.162
I know it's weird when
33:55.176 --> 33:58.838
I do claim that for somebody who is kind of using an off the shelf
33:58.874 --> 34:02.662
thing that they've trained just a little differently, I will put the actual
34:02.796 --> 34:06.550
machine learning pieces in a dependent claim. I try
34:06.600 --> 34:10.126
really hard to keep it out and sometimes I just explain it in
34:10.128 --> 34:12.660
the specification and leave it at that. Right.
34:13.350 --> 34:17.086
Just because it really wasn't the crux of their invention. Right,
34:17.148 --> 34:17.760
right.
34:21.390 --> 34:24.550
You guys thoughts on this stuff? This is just two things I've been
34:24.600 --> 34:27.778
thinking about recently. Is one, should there
34:27.804 --> 34:29.890
be like a specific software patent?
34:31.590 --> 34:35.102
So in the world of utility patent or like patents
34:35.126 --> 34:38.422
and gentlemen, in the US, for example, there would be a design patent, there would
34:38.436 --> 34:42.302
be a utility patent that covered biotech and diagnostics
34:42.386 --> 34:46.080
and that kind of stuff. And it gets a little weird with
34:47.370 --> 34:50.638
digital health and stuff like that. Right. But then
34:50.664 --> 34:54.386
it should be like there's some other kind of software patent where you don't
34:54.458 --> 34:58.162
have to worry about 101 then, because that's the whole reason for that patent class.
34:58.236 --> 35:01.978
If it's for software patenting and maybe it has a
35:02.004 --> 35:05.818
shorter time frame and maybe it's renewable after some time
35:05.844 --> 35:09.610
period, that's the other thing. Software is so quickly. Iterative that
35:09.660 --> 35:12.958
I can imagine a lot of software patents could be obsolete in ten years
35:12.984 --> 35:15.780
just because software has expanded beyond that.
35:17.070 --> 35:20.702
And then could you link to some kind of like code repository
35:20.786 --> 35:24.240
or code listing if it were open source to further enable it?
35:25.530 --> 35:28.810
I don't know, just something because it would obviously remove all the section 101 issues
35:28.860 --> 35:31.558
because that would be the whole purpose of that patent. So it'd really be just
35:31.584 --> 35:34.334
down to novelty, obviousness disclosure,
35:34.382 --> 35:37.918
enablement and things like that. So that's one point for discussion. And the other
35:38.004 --> 35:41.918
point is I think the unfortunate
35:41.954 --> 35:46.118
part of the patent system is that a lot of small inventors
35:46.274 --> 35:49.402
don't it's really hard for them to even recoup their cost of
35:49.416 --> 35:52.898
getting patent. And they may be super innovative, but then the bigger companies do tend
35:52.934 --> 35:56.450
to just squash them and suffocate them in fees
35:56.510 --> 35:58.800
and courts and IPRs and all that kind of stuff.
36:00.210 --> 36:03.758
Recently just heard like, I didn't know much about this, but we were listening
36:03.794 --> 36:07.438
to this werewolf song on the way back from a cider mill this
36:07.464 --> 36:10.994
weekend. And it's the exact same opening
36:11.042 --> 36:14.434
riff as Kid Rock. It was one of Kid Rock's songs. But of course
36:14.472 --> 36:18.310
this werewolf song was like crazy old and Kid Rock song was
36:18.360 --> 36:21.670
not crazy old. We were sitting in the vehicle thinking,
36:21.720 --> 36:25.294
how did Kid Rock, did he just take this and use
36:25.332 --> 36:29.450
it and say, Screw you earlier song? I'm just going to use your Copyright song?
36:29.570 --> 36:33.202
Or did he actually be responsible and pay royalties and
36:33.216 --> 36:36.418
everything? Well, what came out of our little research in the vehicle was
36:36.444 --> 36:39.646
that the Copyright world, especially for music.
36:39.708 --> 36:43.270
I don't know if this transcends into other types of Copyright, but they have
36:43.320 --> 36:46.390
a statutory mechanical royalty rate
36:46.560 --> 36:50.750
that is set forth by the Copyright arbitration loyalty panel
36:50.870 --> 36:54.562
for compulsory mechanical licenses. So basically,
36:54.636 --> 36:58.498
assuming that the work has been previously released to the public, this is
36:58.524 --> 37:02.486
a licensing fee the licensee can pay to sell a cover version
37:02.558 --> 37:06.442
of a song without having to obtain direct permission from the
37:06.456 --> 37:10.090
rights holder. So in the US, this rate is 9.1 cents per track,
37:10.140 --> 37:13.534
or 1.5 cents for each minute of playing time,
37:13.692 --> 37:17.050
whichever is greater. So I'm just wondering, to make it more fair
37:17.100 --> 37:21.300
for smaller inventors, smaller companies,
37:21.630 --> 37:25.522
is there a world where we need to have some kind of like statutory rate
37:25.656 --> 37:29.266
where if we assume all patents are valid because they did go through this crazy
37:29.448 --> 37:33.394
patent office procedure before
37:33.432 --> 37:36.862
you release the product, you go through some kind of diligence. And if
37:36.876 --> 37:40.740
you find patents that reasonably cover what you're doing,
37:41.250 --> 37:44.498
even if you can maybe make an argument that the one word in the patent
37:44.534 --> 37:48.274
that you may or may not do is there some base
37:48.432 --> 37:51.802
rate that you have to pay to that inventor? Because it
37:51.816 --> 37:55.322
assumes that the Usto job does job correctly, it assumes the patent
37:55.346 --> 37:58.870
is bad, assumes everybody was good actors, and it allows for at least
37:58.920 --> 38:02.242
some cost to be recuperated by
38:02.256 --> 38:05.614
full inventors. I don't know, just weird things I've been thinking about
38:05.652 --> 38:09.478
lately. So curious about your thoughts on the whole software patent idea and
38:09.504 --> 38:13.310
the statutory rate idea for patents, the software
38:13.490 --> 38:16.426
like patent Copyright sort of idea.
38:16.548 --> 38:19.934
It just reminds me of. And this is a really odd
38:19.982 --> 38:23.710
thing. I haven't dealt with this a lot, but I know in the semiconductor world
38:23.760 --> 38:25.680
there's something called a mask work.
38:29.530 --> 38:33.354
Intel designs a processor, and it could have hundreds
38:33.402 --> 38:36.422
steps in the process. And each one of those is a pattern that needs a
38:36.436 --> 38:40.218
mask. It all fits together in this big 3D layout. And it's
38:40.254 --> 38:44.382
like, sure, there's lots of patentable technology in there, but it's
38:44.406 --> 38:48.434
almost like a whole separate thing. How do you get everything to
38:48.472 --> 38:51.554
fit on one chip? How big do you make
38:51.592 --> 38:54.278
this line? How big do you make that line? What do you put next to
38:54.304 --> 38:57.606
each other? And there's a lot of sort of design innovation
38:57.678 --> 39:01.022
whatever in that. So it's basically just a
39:01.036 --> 39:04.434
Copyright, though. It's not even a design patent
39:04.482 --> 39:07.634
where there's like a claim, what we show
39:07.672 --> 39:11.622
in these figures or whatever. You can word that in design path in a mask
39:11.646 --> 39:14.570
work, apparently it's just like a Copyright but it's,
39:15.310 --> 39:18.738
I guess a vehicle for that, for a specific mask
39:18.774 --> 39:23.080
set in the semiconductor world. I wonder if there could be a special
39:24.670 --> 39:27.938
software Copyright or something that would do
39:27.964 --> 39:31.346
something similar. Yeah, no, I like that. I think it's really interesting,
39:31.468 --> 39:35.402
I think, pulling from these different spaces because I think there is
39:35.416 --> 39:39.062
an issue with software pens. And I know there's been a
39:39.076 --> 39:42.566
lot of success, but I think there could be something better or at least
39:42.628 --> 39:45.758
like parallel, another track for people to
39:45.784 --> 39:48.902
take. So I think that's why I want to look into that more and see
39:48.916 --> 39:52.646
how that's done, because I think that's really interesting. So I'm a huge
39:52.708 --> 39:54.280
dissenter of this idea.
39:58.970 --> 40:03.298
Most of it is this. I've written code, I've designed
40:03.334 --> 40:06.762
consumer products myself. And then I've moved into the patent world and
40:06.776 --> 40:09.966
done everything that you could possibly do with software there.
40:10.088 --> 40:13.220
And this is my argument. This phone,
40:13.850 --> 40:17.310
this is a paperweight without software.
40:18.590 --> 40:22.580
It's a piece of metal. It's a chunk of nothing.
40:23.150 --> 40:26.190
It needs the software to do what it does. That's a utility.
40:27.470 --> 40:30.738
If you then move and you say, okay, let's go online and
40:30.764 --> 40:34.734
talk about some HTML code that sends me an email
40:34.892 --> 40:40.090
that is no longer a definitive
40:40.150 --> 40:43.938
device. It isn't like a widget as a mechanical design.
40:44.024 --> 40:45.990
Right. Or a mechanical patent.
40:47.390 --> 40:51.738
It's kind of in the ether. Right, right. But the
40:51.824 --> 40:55.580
engineering behind that has an algorithm and has a design,
40:56.030 --> 40:59.734
and it took an input, it created an output
40:59.842 --> 41:03.678
and it created a display of an output. So those really are the
41:03.704 --> 41:07.806
three things that you typically do with anything
41:07.928 --> 41:11.178
outside of mechanical. Mechanical could just be
41:11.204 --> 41:14.514
a widget. Right. But almost every other
41:14.552 --> 41:18.162
piece of a patent would be take an input, do something, provide an
41:18.176 --> 41:20.600
output. And that's what software does.
41:21.050 --> 41:24.298
And most of our devices without software are just bricks.
41:24.454 --> 41:27.498
Right. Falls apart a little bit for me.
41:27.524 --> 41:31.134
But Unfortunately, I don't think the courts are
41:31.172 --> 41:35.070
viewing that's where there was one of the cases
41:35.570 --> 41:39.346
that it was a camera, like the device was directed
41:39.358 --> 41:42.870
to a camera, but it was like a way to get better,
41:42.920 --> 41:46.702
CRISPR images with a camera or something like that. And it was held
41:46.726 --> 41:50.130
Invalid by either the district court. I think it was a federal circuit.
41:50.870 --> 41:54.106
To me, that's like so far beyond aspect
41:54.178 --> 41:55.890
idea because it's a camera.
41:57.470 --> 42:01.234
And again, I think there's some conflagration
42:01.342 --> 42:04.998
of 103, which is causing some of these
42:05.024 --> 42:08.382
issues as well. And like enablement and sufficiency of disclosure, all that
42:08.396 --> 42:12.442
kind of stuff. Sure. But they're
42:12.466 --> 42:15.762
not fixing it in that way like they should. So I guess this
42:15.776 --> 42:19.160
is like an end around. If you're going to fix it like you should
42:19.670 --> 42:22.986
do, we just go around you and do something
42:23.048 --> 42:26.574
different. I understand. I think
42:26.612 --> 42:30.238
there's like a lot of issues with Alice, with how it conflates 102
42:30.283 --> 42:33.174
and 103 with 101 and all those things.
42:33.212 --> 42:37.078
But Kristen Ashley, I totally
42:37.114 --> 42:40.666
agree. And like there's a lot of invention innovation
42:40.738 --> 42:44.238
that happens when you write code. And so I
42:44.264 --> 42:47.406
think part of AI is a great example,
42:47.468 --> 42:51.246
too. I think most people,
42:51.428 --> 42:54.834
the examiner, the inventor, the agent,
42:54.932 --> 42:59.038
attorney would all agree there is something here that's been invented.
42:59.194 --> 43:02.622
But I think where everybody gets hung up, at least the examiner and
43:02.636 --> 43:05.840
the agent or attorney is in the claim language itself.
43:06.650 --> 43:10.506
How did you write it down? So that it is
43:10.688 --> 43:14.382
only covering the real thing that you invented and not
43:14.516 --> 43:18.294
other crazy abstract ideas or other things that are out there or whatever.
43:18.392 --> 43:21.438
And that's where I think in the semiconductor world, again,
43:21.584 --> 43:25.170
I'm much more on the material side and device side.
43:25.220 --> 43:29.082
But in the semiconductor world, there's a lot of stuff when you design a
43:29.096 --> 43:32.614
chip that is patentable, but then there's this whole other component
43:32.662 --> 43:35.240
of it, which is how does everything fit together,
43:36.050 --> 43:39.526
which is sort of how do you write that in a claim?
43:39.598 --> 43:43.146
God, I mean, it would be like thousands and thousands of lines long
43:43.208 --> 43:46.078
to capture all of those design intricacies.
43:46.234 --> 43:48.620
So maybe a situation where there's like,
43:49.130 --> 43:52.746
value for both sorts of things, there's value
43:52.808 --> 43:56.446
in figuring out what is the crux
43:56.518 --> 44:00.282
of the innovation that you've done here with this code that no one
44:00.296 --> 44:03.882
else has been able to do or take this kind of input and display it
44:03.896 --> 44:07.086
in a certain way or process it differently or whatever. There also
44:07.148 --> 44:10.350
could be value in somebody else. It might have taken
44:10.400 --> 44:13.938
them 20,000 lines to write this code. I did it in 4000
44:14.084 --> 44:18.406
because I figured out a very efficient
44:18.478 --> 44:22.218
way to organize everything or whatever. And that might be where like
44:22.244 --> 44:25.722
a mask work or a software Copyright in a
44:25.736 --> 44:29.180
way like could be a complimentary, right?
44:29.510 --> 44:32.250
Yes. I like this as an alternative.
44:33.350 --> 44:36.226
I don't like it on its own, but I do like it as an alternative,
44:36.298 --> 44:39.894
especially with somebody who either doesn't want to pay to argue over
44:39.932 --> 44:43.134
Alice or doesn't know how to argue over
44:43.172 --> 44:47.082
Alice. Right. Or doesn't have enough in their specification to
44:47.096 --> 44:50.562
get them anywhere. Right. And to your point
44:50.696 --> 44:54.378
about really lengthy claims, that is how
44:54.404 --> 44:57.462
we got over Alice when it first came out the first year or two,
44:57.596 --> 45:00.270
you just kept adding on another dongle.
45:01.190 --> 45:04.760
It was awful. And you would get the worst claims because
45:05.510 --> 45:08.526
you didn't want to get the client nothing. I did a lot
45:08.528 --> 45:11.802
of work for SAP back then and that's business method stuff,
45:11.876 --> 45:15.850
and they were attacked horribly every time they went anywhere
45:15.970 --> 45:17.120
with a claim set.
45:19.650 --> 45:23.002
That's interesting. Let me bump you
45:23.016 --> 45:25.990
all. For example, I've been really thinking about this, and I think there's an opportunity
45:26.100 --> 45:29.770
for some kind of bigger discussion about just
45:29.820 --> 45:33.274
alternatives, like even just to put my mind
45:33.312 --> 45:37.140
at rest at night about these things.
45:38.190 --> 45:40.894
What are people's thoughts on this? And I know we have to probably get going
45:40.932 --> 45:45.060
here, but anybody's thoughts on a statutory rate for
45:45.510 --> 45:49.198
the patent world. I don't know how it would be policed or
45:49.284 --> 45:52.762
set up or whatever. I think
45:52.776 --> 45:56.482
it's interesting. It's tough. The one
45:56.556 --> 45:59.962
challenge might be beyond something like in the music industry. And a song is a
45:59.976 --> 46:03.158
song. It kind of has a standard scope,
46:03.254 --> 46:07.258
if you will. Certain markets can go into whatever with
46:07.284 --> 46:11.662
a lot of software and things like that. How do you decide on is
46:11.676 --> 46:15.298
it the same rate for a little piece
46:15.324 --> 46:19.154
of software in that sensor of a little photo detector,
46:19.202 --> 46:22.702
some super simple thing versus like a computer or an
46:22.716 --> 46:24.960
operating system or something? Yeah.
46:25.710 --> 46:28.882
So a piece of that that you have to remember when talking to any of
46:28.896 --> 46:32.770
your software clients. If somebody comes to you with something that works
46:32.820 --> 46:36.118
way in the back end of the computer and you cannot detect it,
46:36.264 --> 46:39.634
that's a terrible claim. You don't even want to go
46:39.672 --> 46:43.630
after that. That's just somebody who needs to bulk up their portfolio. It's really
46:43.680 --> 46:47.066
useless to have a non detectable claim because who's going to infringe and how you're
46:47.078 --> 46:50.590
going to find it, right? Yeah. I just don't know how to solve the problems.
46:50.640 --> 46:53.738
There's just so many small inventors,
46:53.774 --> 46:57.218
small companies that develop really great stuff and other bigger
46:57.254 --> 47:01.538
companies that just can move way faster, have bigger budgets, have bigger
47:01.574 --> 47:05.350
purses, just pick it up and do it and then
47:05.460 --> 47:08.734
smolder them with IPRs and
47:08.772 --> 47:12.118
court cases. I know AI was supposed to fix some
47:12.144 --> 47:14.950
of that, but I think it just made it worse. And then I was just
47:15.000 --> 47:18.362
reading last night some of the newer reforms
47:18.386 --> 47:22.510
that lady and the other one are suggesting.
47:23.190 --> 47:26.482
In my opinion, it would actually make it even worse where it
47:26.496 --> 47:30.314
would be even easier to Institute PPAP receivings
47:30.362 --> 47:33.542
and it would be even easier to cancel claims,
47:33.686 --> 47:37.498
all these things. That would make it even harder, I think, for smaller inventors to
47:37.524 --> 47:41.146
really protect themselves. I definitely get a
47:41.148 --> 47:44.266
few phone calls a year, at least, if not more,
47:44.328 --> 47:47.990
where you have these inventors that really seem to have a legit
47:48.110 --> 47:51.382
thing. And I connected with people that do litigation, things like
47:51.396 --> 47:55.774
that. At the end of the day, nobody wants to pick up litigation on
47:55.812 --> 47:59.050
a like if we win kind of basis,
47:59.730 --> 48:04.214
then you have to pay us. But if not contingency
48:04.262 --> 48:07.942
basis. There we go. That's the word. What's the
48:07.956 --> 48:11.482
remedy for the small guys Anywho? Well, thanks for
48:11.496 --> 48:12.840
entertaining me all today.
48:15.370 --> 48:18.590
Yeah. Thanks, Ashley. It's a great topic. If you have anything
48:18.640 --> 48:21.938
else you want to talk about. Add, I think you should write an
48:21.964 --> 48:26.380
article about this. Actually, I think we're thinking about doing because
48:27.910 --> 48:30.938
I've been doing more bigger thinking.
48:31.084 --> 48:34.538
I don't know about things and it's things that keep me up at
48:34.564 --> 48:38.246
night, literally, but putting something together.
48:38.308 --> 48:41.702
I made the same thing. Like, Kristen, if you have some thoughts, if you got
48:41.716 --> 48:45.026
some thoughts, let me know because I might all do one,
48:45.088 --> 48:50.278
like by everybody article and everybody porting
48:50.314 --> 48:53.518
over some of these other concepts into patent
48:53.554 --> 48:56.794
laws because some of the stuff that they're suggesting
48:56.842 --> 49:00.378
right now in Congress, I'm like dear Lord, you all have
49:00.404 --> 49:04.220
no idea what pudding you're putting your finger into
49:04.910 --> 49:07.326
and how you're messing it up.
49:07.448 --> 49:10.686
So maybe there's nothing but at least
49:10.748 --> 49:12.440
it helps me sleep better at night.
49:16.670 --> 49:20.314
All right, well, thank you. Thank you, Ashley.
49:20.422 --> 49:22.962
Thank you. Nice to meet everybody. Thanks.
49:23.036 --> 49:26.070
Bye. Bye. Bye. Bye. All right.
49:26.120 --> 49:30.474
That's all for today, folks. Thanks for listening. And remember to check us out@aurorapatons.com
49:30.512 --> 49:33.994
for more great podcasts, blogs and videos covering all things patent
49:34.042 --> 49:37.062
strategy and if you're an agent or attorney and would like to be part of
49:37.076 --> 49:40.878
the discussion or an inventor with a topic you'd like to hear discussed, email us
49:40.904 --> 49:44.298
at podcast at Aurora patents.com, do remember that this
49:44.324 --> 49:47.838
podcast does not constitute legal advice and until next time keep calm and
49:47.864 --> 49:48.420
patent on.
