[00:00:00] Speaker 01:
So we will hear argument next in case number 232323, Swarm Technologies against Amazon.

[00:00:11] Speaker 01:
Ms.

[00:00:11] Speaker 01:
Addy.

[00:00:12] Speaker 02:
Good morning, Your Honors.

[00:00:13] Speaker 02:
May it please the court, I would like to start with the 101 issue here.

[00:00:20] Speaker 02:
The board's decision must be reversed because it is contrary to Alice and Enfish.

[00:00:27] Speaker 02:
These are hardware claims.

[00:00:30] Speaker 02:
And they are configured architecture that improves the operation of the computer itself, not just a known process that's run on a general purpose computer.

[00:00:41] Speaker 02:
And because of that, these claims fit squarely within the Enfish safe harbor.

[00:00:47] Speaker 01:
Are we talking about the substitute claims 13 and 14?

[00:00:51] Speaker 02:
Yes, Your Honor.

[00:00:52] Speaker 01:
And do I remember right?

[00:00:54] Speaker 01:
The board said it was going to reject those on two independent grounds.

[00:00:59] Speaker 01:
One is 101, and the other is written description support.

[00:01:02] Speaker 02:
That's correct, Your Honor.

[00:01:03] Speaker 02:
And I would like to address 112 after 101, if that is OK with the court.

[00:01:09] Speaker 02:
Sure.

[00:01:10] Speaker 02:
So Your Honors, these claims fit right into this in Fish Safe Harbor, because they are improvements to computer architecture and configuration to computer itself.

[00:01:22] Speaker 02:
Let's look at what the claims are directed to.

[00:01:25] Speaker 02:
Claim 13 is directed to component architecture, where the controller populates the task pool and the co-processors go to the task pool autonomously and proactively and search for tasks to perform.

[00:01:41] Speaker 02:
This is different from conventional systems, which were known as controller responder systems.

[00:01:46] Speaker 03:
Do I understand correctly that what the board was saying is that you've got these, what's it called, a scrum board, right?

[00:01:53] Speaker 03:
And they're saying there's been a long-known conventional way of organizing human activity using these scrum boards.

[00:02:02] Speaker 03:
And so the abstract idea is organizing computer activity, I guess, using a well-known way of organizing human activity.

[00:02:13] Speaker 03:
That's the way I've been kind of thinking about what they

[00:02:16] Speaker 03:
were saying in terms of their analysis and saying that that's an abstract idea.

[00:02:21] Speaker 02:
Well, two points.

[00:02:22] Speaker 03:
Do you just agree that that's what they said?

[00:02:24] Speaker 02:
Not exactly, Your Honor.

[00:02:26] Speaker 02:
I agree that they said that the extracted limitations of the claim to recite the concept of a scrum board.

[00:02:35] Speaker 02:
And then they said that that could be performed by humans.

[00:02:39] Speaker 02:
But when you look at the extracted limitations of the claims,

[00:02:44] Speaker 02:
That's not all of the claims.

[00:02:45] Speaker 02:
And you can see that on page 91 is where they say the extracted elements, 92.

[00:02:50] Speaker 02:
And if you go back to 91, you can see that they're only taking out the functional limitations in the claim and saying they recite the concept of a scrum board.

[00:02:59] Speaker 02:
And that's contrary to Alice's directive.

[00:03:01] Speaker 03:
I thought that what they were saying was that if you think of the controller, I suppose, as a manager, and you think of the co-processors as workers, that it's analogous to the idea of having humans having a scrum board where the worker walks up and grabs a task.

[00:03:21] Speaker 03:
and then performs it and then puts it back.

[00:03:25] Speaker 03:
And the managers, they're seeing that the tasks are being done and adding tasks to the Scrum Board.

[00:03:31] Speaker 03:
I'm not saying that that's ineligible.

[00:03:33] Speaker 03:
I'm just asking if you agree that that's the analogy that the board was looking at.

[00:03:38] Speaker 03:
And then the question becomes whether, sorry, you go ahead and answer that question.

[00:03:42] Speaker 02:
I'm sorry, Your Honor.

[00:03:43] Speaker 02:
That is the second problem with the boards directed to analysis.

[00:03:49] Speaker 02:
After extracting certain limitations, then it changed other limitations.

[00:03:53] Speaker 02:
It changed the structural limitations of a controller, a coprocessor, and a task pool to a worker and a manager and a thrombord.

[00:04:01] Speaker 02:
But the claims have nothing to do with that.

[00:04:04] Speaker 02:
And Your Honor, you mentioned an analogy, but it's not enough to simply identify some real world analogy.

[00:04:12] Speaker 02:
Alice requires much more rigor that you look at what the claims are directed to,

[00:04:16] Speaker 02:
not what they're analogous to or, as the board said, what they're configured to.

[00:04:24] Speaker 02:
In fact, this court has addressed that analogous to issue because it's come up before in Data Engine.

[00:04:30] Speaker 02:
And in Data Engine, this court said it's not enough to merely identify a patent eligible concept in the claims.

[00:04:37] Speaker 02:
You have to be rigorous and look at the claim language itself

[00:04:40] Speaker 02:
but individually as a whole, and that the board did not do here.

[00:04:44] Speaker 00:
Do you view the board as having rejected as a factual matter that your claims are directed to improving computer functionality, or do you view them as saying, OK, we'll accept that they improve computer functionality, but because the same idea maybe abstractly would help in other situations, it doesn't count?

[00:05:05] Speaker 02:
Your Honor, I don't believe it was a factual rejection, but I believe that because the board

[00:05:09] Speaker 02:
extracted some elements and changed other elements that it failed to look at the claim language as a whole and the spec, which requires the elements to be configured specifically.

[00:05:26] Speaker 02:
I'm sorry, did I answer your question?

[00:05:27] Speaker 02:
Can everyone make sure I covered it?

[00:05:31] Speaker 02:
So, Your Honors, Data Engine has already dealt with this analogy.

[00:05:36] Speaker 02:
theory and data engines dealt with the configure to theory.

[00:05:41] Speaker 02:
And the board simply did not provide a directed to analysis under step one that looks at the claims both individually and the language as a whole.

[00:05:52] Speaker 02:
Because if it did, it would have found that the claims were patent eligible because they fit that

[00:05:58] Speaker 02:
in Fish Safe Harbor of improving computer functionality.

[00:06:02] Speaker 02:
I want to address my question, your question one more time.

[00:06:04] Speaker 00:
I think you, if my colleagues don't mind, I'm curious about written description.

[00:06:09] Speaker 00:
As I understand it here, it's the without any communication limitation and here the issue is whether that's broad enough

[00:06:21] Speaker 00:
to include within the language both indirect and direct communication.

[00:06:25] Speaker 00:
So therefore, because it's a negative limitation, you need written description support for having no direct and no indirect communication.

[00:06:37] Speaker 00:
Did I say that right?

[00:06:38] Speaker 02:
That's the way the board looked at it.

[00:06:40] Speaker 02:
But Your Honor, the board, on its 112 analysis, construed without any communication with the controller to be

[00:06:47] Speaker 02:
contrary to the plain plain language and to be contrary to the construction of that same limitation that it used in applying its 103 analysis.

[00:06:58] Speaker 02:
Your honors, in the 103 analysis, the 103 construction is correct.

[00:07:03] Speaker 02:
Amazon proposed that construction we did not dispute it and the board adopted it Yes Amazon proposed the construction first where in the board opinion is that it's a 103 construction Yes, your honor it's at I

[00:07:32] Speaker 02:
One moment, your honor.

[00:07:37] Speaker 02:
So you're looking for the board's construction on obviousness.

[00:07:42] Speaker 00:
Right.

[00:07:43] Speaker 00:
The construction of any communication, how they understood the term.

[00:07:46] Speaker 02:
Can I walk you through how they got there?

[00:07:49] Speaker 02:
And then I can get to the... I don't think it's stated, this is the construction, but I think it's required.

[00:07:58] Speaker 02:
And that's because in the petition,

[00:08:01] Speaker 02:
Amazon said without any communication with the controller occurs because the co-processors only communicate with the intermediary.

[00:08:09] Speaker 01:
And then they said that because of that... Do you have an appendix page for that?

[00:08:15] Speaker 02:
Yes, appendix page 383.

[00:08:23] Speaker 02:
And because of that, your honors, dynamic acceptance, according to Amazon, occurs without communicating with the TCC in Laurie, which it equivalizes to the coprocessor, to the controller, because the TCW contacts the intermediary.

[00:08:39] Speaker 02:
In Laurie, that's the ATD mechanism.

[00:08:41] Speaker 02:
So let me say that again.

[00:08:42] Speaker 02:
Dynamic acceptance of a coprocessor occurs without communicating with the TCC,

[00:08:49] Speaker 02:
because the TCW contacts the intermediary, hear the ATD mechanism instead.

[00:08:56] Speaker 02:
That's at 383.

[00:08:58] Speaker 00:
And then- Sorry.

[00:08:58] Speaker 00:
So the petition was allowing for indirect communication to occur both in your claims and therefore thought they could invalidate your claims by pointing to Lurie having indirect communication.

[00:09:11] Speaker 00:
Is that what you're saying?

[00:09:12] Speaker 02:
Yes.

[00:09:13] Speaker 02:
And I think indirect communication should also be understood to mean

[00:09:18] Speaker 02:
communication with the intermediary, because it's not communication with the controller.

[00:09:24] Speaker 02:
It's communication just with the intermediary.

[00:09:26] Speaker 02:
And the board called it indirect communication.

[00:09:29] Speaker 02:
Sometimes it comes up in the spec as that as well.

[00:09:33] Speaker 02:
Your Honors, I'm sorry I didn't have that cite for you earlier.

[00:09:35] Speaker 02:
The decision, the place where the board applied its obviousness decision is at appendix page 40.

[00:09:42] Speaker 02:
And the board said, citing Lurie at 1251, as such,

[00:09:50] Speaker 02:
the ATD mechanism provides a level of indirection between the TCC and the TCW.

[00:09:58] Speaker 02:
So the board applied the construction wherein that level of indirection, that communication with the intermediary is not included in the language without any communication.

[00:10:14] Speaker 02:
But then, when the board got to 112, it did the exact opposite.

[00:10:20] Speaker 02:
because those two constructions cannot be reconciled.

[00:10:25] Speaker 02:
And if you look at the claim language, the claim language itself is replete with this indirect communication where the co-processors only communicate with the task pool, not with the controller.

[00:10:39] Speaker 02:
For example, we're looking here at 13G, but if you also look at 13C and 13D, they talk about the activities that the co-processors perform in searching for

[00:10:50] Speaker 02:
and receiving the task from the task pool.

[00:10:55] Speaker 02:
And they do all of that without, quote, all without any communication between the co-processor and the controller.

[00:11:03] Speaker 02:
So this indirect communication lives in harmony with the without any communication with the controller.

[00:11:12] Speaker 03:
Turning to the written description, where do you think your best support is for the without any communication limitation?

[00:11:20] Speaker 02:
So Your Honor, the specification talks about it two ways.

[00:11:25] Speaker 02:
First, the specification is clear that the co-processors communicate with the task pool.

[00:11:30] Speaker 02:
And that's the point of this invention, because as we were discussing in the prior argument, relieves burden on the controller.

[00:11:38] Speaker 02:
You can see that at appendix 151, lines

[00:11:42] Speaker 02:
Column 2, lines 14 and 19.

[00:11:45] Speaker 03:
Is that in connection with obtaining tasks and performing tasks, or is that in connection with the plug and play of the processor?

[00:11:54] Speaker 03:
It's both, Your Honor.

[00:11:56] Speaker 03:
OK, so where does it talk about no communication with respect to the plug and play?

[00:12:14] Speaker 03:
or just communication with the task pool with respect to the plug and play.

[00:12:22] Speaker 02:
Your Honor, I believe that that is at column eight to nine.

[00:12:31] Speaker 02:
No, that's not.

[00:12:37] Speaker 02:
Here we go.

[00:12:38] Speaker 02:
I'm sorry, Your Honors.

[00:12:45] Speaker 02:
at column 10, line 29, but moving on through that paragraph to about line 32, it says, moreover, the task pool itself may exclude particular cells on the basis of low performance, unreliable connection, or poor data throughput.

[00:13:06] Speaker 03:
And it also says- How is that plug and play?

[00:13:08] Speaker 03:
I have one more.

[00:13:09] Speaker 03:
I'm sorry.

[00:13:10] Speaker 02:
Oh, I'm sorry.

[00:13:10] Speaker 03:
I just want to know how that is in your mind.

[00:13:13] Speaker 03:
How do I know that's plug and play?

[00:13:20] Speaker 02:
Well, it is plug and play because one of the advantages to the system is that the system can use the coprocessing capabilities of other coprocessors that aren't already connected to the system but are on the same network.

[00:13:43] Speaker 02:
And because of that advantage, that is how

[00:13:48] Speaker 02:
the plug and play works and there's an example specifically in the patent about a smart light bulb where that smart light bulb is in the network but it's not currently connected to the system and the smart light bulb due to its programming and configuration recognizes that the task pool is overloaded and that smart light bulb then joins the system to help

[00:14:12] Speaker 02:
process the tasks required by the task pool.

[00:14:15] Speaker 03:
Okay, I interrupted you.

[00:14:16] Speaker 03:
Did you want to give us the other site?

[00:14:20] Speaker 02:
I want to use the example of the smart light bulb, and I'm looking for that site right now.

[00:14:29] Speaker 02:
Okay, a couple of sites.

[00:14:35] Speaker 02:
Let's see.

[00:14:38] Speaker 02:
First of all, at page 155,

[00:14:42] Speaker 02:
I think we just talked about this.

[00:14:44] Speaker 02:
Column 10, lines 2 through 34, the task pool itself may exclude particular cells.

[00:14:51] Speaker 02:
And page 155, column 10, 45 to 47, the co-processors may connect to the task pool, such as being by wired configuration, which I'm sorry doesn't relate to the plug and play, but certainly relates to no communication with

[00:15:07] Speaker 02:
the controller because they can't communicate with the controller if they're wired to the task pool.

[00:15:14] Speaker 00:
Let me quickly take you back to the construction.

[00:15:16] Speaker 00:
Did you originally have a claim, I think it was claim 20, that expressed the call out without direct communications with the CPU?

[00:15:24] Speaker 00:
And if you did, doesn't that suggest the patentee had in mind that

[00:15:31] Speaker 00:
a distinction between direct and indirect.

[00:15:33] Speaker 00:
And so when they said without any communication, they must demand to exclude all types of communication.

[00:15:39] Speaker 02:
Your Honor, I believe it's claim 10.

[00:15:41] Speaker 02:
And claim 10 says that the process can be performed without any indirect communication with the controller.

[00:15:47] Speaker 02:
So claim 10 fits nicely within this paradigm we're discussing where this communication with the task pool, which they're calling indirect communication with the controller, doesn't satisfy

[00:16:00] Speaker 02:
communication with the controller.

[00:16:05] Speaker 02:
Your honors, I see I'm into my rebuttal time.

[00:16:08] Speaker 02:
I would address 103 or I can... Yes, your honors.

[00:16:26] Speaker 04:
Good morning, and may it please the court.

[00:16:28] Speaker 04:
My name is Adam Greenfield on behalf of Amazon.

[00:16:30] Speaker 04:
To reset, there's two separate appeals here.

[00:16:33] Speaker 04:
The first, regarding the 04 patent, the board agreed with Amazon that all claims would have been obvious and that the proposed substitute claims are not patentable.

[00:16:43] Speaker 04:
The court should affirm that decision.

[00:16:45] Speaker 04:
Separately, there's the 275 patent, and there the board upheld the claims based on one limitation,

[00:16:53] Speaker 04:
That is, whether co-processors can proactively retrieve a task from the task pool.

[00:16:59] Speaker 04:
Amazon appeals that decision because it's based on an erroneously narrow construction that excludes express embodiments in the specification and violates the law of obviousness.

[00:17:09] Speaker 01:
Can you start by talking about 004 Claim 13 and the written description point?

[00:17:15] Speaker 04:
Absolutely.

[00:17:17] Speaker 04:
The written description

[00:17:18] Speaker 04:
claim 13, oh, the proposed substitute claim.

[00:17:21] Speaker 04:
Correct, Your Honor.

[00:17:22] Speaker 04:
All of the examples that SWARM has pointed to don't relate to dynamic acceptance of co-processors or plug and play.

[00:17:30] Speaker 04:
They relate to different aspects of the claim.

[00:17:33] Speaker 04:
And here it's their burden to show that there is support, that there be an expressed disclosure in the 332 application, not the specification, the 332 application.

[00:17:43] Speaker 04:
And the board agreed with us that they failed to do so.

[00:17:46] Speaker 04:
to, Your Honor, Judge Stark's point.

[00:17:51] Speaker 04:
You're right, in the original application, there is, I think it's claim 20 that talks about direct communication.

[00:17:55] Speaker 04:
That's a totally separate issue than the issue of claim 10 that I think Swarm was directing you to.

[00:18:02] Speaker 04:
And here, Swarm's position on the communication and how it, without any communication, actually means with only indirect communication,

[00:18:13] Speaker 04:
is really in tension with the argument that they're making in the Juniper Appeal, where they've said that they're brief, in that case at 31 to 32, that this same limitation, dynamic acceptance without any communication, is plain.

[00:18:25] Speaker 04:
There's no definition in the specification.

[00:18:27] Speaker 04:
You don't have to look for extrinsic evidence.

[00:18:29] Speaker 04:
And all of that supports our argument.

[00:18:31] Speaker 04:
And the board's finding that there's no written description support.

[00:18:36] Speaker 00:
What about the argument that the board's understanding, essentially its construction, was different in the 103 and the 112 context?

[00:18:45] Speaker 04:
That's not true, Your Honor.

[00:18:47] Speaker 04:
So first, this limitation in the 103 context wasn't disputed.

[00:18:54] Speaker 04:
And so the board's analysis of that also comes up in the context of claim one, which Swarm hasn't appealed.

[00:19:00] Speaker 04:
So just to reset.

[00:19:02] Speaker 04:
We, uh, the board found claim one, which includes this limitation would have been obvious.

[00:19:06] Speaker 04:
And if this specific limitation wasn't disputed by swarm and swarm hasn't appealed that.

[00:19:11] Speaker 00:
But the board did go, I know it's not on appeal, but the board did go on to find that Lori, I think it helped me if I'm wrong, disclosed, uh, without any communication, even though Lori seems to have indirect communication.

[00:19:26] Speaker 04:
not with respect to this limitation, the dynamic acceptance.

[00:19:30] Speaker 04:
All Lori says is that sort of at an abstract level, the ATD mechanism goes in between the TCC and the TCW, so it's an intermediary.

[00:19:42] Speaker 04:
But it doesn't say that for purposes of dynamic acceptance of TCWs, there is any indirect communication.

[00:19:49] Speaker 04:
We didn't argue that.

[00:19:50] Speaker 04:
In fact, we pointed to the activation demon that the board relied on.

[00:19:53] Speaker 04:
and that doesn't involve any communication back to the TCC.

[00:19:57] Speaker 04:
And I would just, to the extent that you look at our arguments in the petition that they're pointing to, look very carefully.

[00:20:04] Speaker 04:
None of the statements they're pointing to is Amazon arguing that Laurie

[00:20:10] Speaker 04:
TCWs indirectly communicate with the TCC when adding TCWs.

[00:20:14] Speaker 04:
That's not something we argued at all.

[00:20:15] Speaker 04:
It's sort of more general references to the ATD as an intermediary generally for certain things between the TCC and the TCW.

[00:20:30] Speaker 04:
Responding to the point on the 04, Swarm pointed you to A40 of the board's decision.

[00:20:37] Speaker 04:
I think if you take a closer look, it relates to a different element.

[00:20:42] Speaker 04:
It's not this dynamic acceptance element at all.

[00:20:47] Speaker 04:
And so we agree that with the...

[00:20:50] Speaker 04:
with the board here that everything they've pointed to for a description doesn't relate to the settlement.

[00:20:55] Speaker 04:
I'll move briefly to the 101 issue that they raised, so that I have time on the 275 patent.

[00:21:02] Speaker 04:
We agree with Judge Stoll's framing of the board's 101 decision and how it looked at the abstract idea.

[00:21:08] Speaker 04:
The board, in that sense, agreed with Judge Jimado in Arizona, Judge Donato in Northern District of California when this issue came up.

[00:21:16] Speaker 03:
I appreciate there's a different context.

[00:21:17] Speaker 03:
Isn't this different, though?

[00:21:20] Speaker 03:
You know, a lot of times when you say, oh, this is just organizing human activity on a computer, do it on a computer.

[00:21:26] Speaker 03:
They're talking about something like a business method, you know, or a dating app or bingo or something like that.

[00:21:32] Speaker 03:
Do it on a computer here.

[00:21:34] Speaker 03:
Why isn't this borrowing an idea of how organizing human activity and saying, let's see if this will work on a computer system.

[00:21:44] Speaker 03:
Oh, it works.

[00:21:45] Speaker 03:
It actually makes the computer system work more efficiently.

[00:21:47] Speaker 03:
It's a.

[00:21:49] Speaker 03:
you know, providing, you know, what do they call that, you know, divided tasking or whatever by the computer, right?

[00:21:57] Speaker 03:
And so that it's the same as what's been done on a scrum board, the recognition that something like that might improve a computer network.

[00:22:06] Speaker 03:
Why isn't that a technological solution to a technological problem?

[00:22:10] Speaker 04:
Two things are under it.

[00:22:11] Speaker 04:
The first is all of the, first of all, we said that the technical, the problem here is not unique to technology, this idea of speeding up and parallel processing.

[00:22:21] Speaker 04:
The same would be true for processing tasks in the real world in the Scrum Board example.

[00:22:27] Speaker 04:
And in this case, in this court's decisions in Simeo and PSG, they say the inventive concept can't be inherent in the abstract idea.

[00:22:36] Speaker 04:
And so everything that,

[00:22:38] Speaker 04:
that you reference and that they're pointing to, that is the abstract idea of a scrum board of returning tasks, completing tasks.

[00:22:45] Speaker 03:
But that's a scrum board for humans.

[00:22:47] Speaker 03:
And your view is that the idea of making a computer work better by using something that's made humans work better,

[00:22:56] Speaker 03:
You know, why is that not invented?

[00:22:59] Speaker 04:
That's my question.

[00:23:00] Speaker 04:
They would have to go the next step and have an actual technical solution.

[00:23:04] Speaker 04:
Here, the controller, the task pool, these are all described at a very high level in the specification.

[00:23:12] Speaker 04:
There's no suggestion in the patent that there's any technological changes.

[00:23:16] Speaker 04:
The only thing was the organization, the arrangement of conventional existing components in the manner of a scramble.

[00:23:22] Speaker 03:
Well, what if the arrangement is a way that's never been done before, assuming that?

[00:23:26] Speaker 04:
Right.

[00:23:27] Speaker 04:
Well, the 102, 103 ounce is different.

[00:23:28] Speaker 04:
But here, if that arrangement just mimics the fundamental way of human organization, that's insufficient.

[00:23:35] Speaker 00:
Have we ever said that?

[00:23:36] Speaker 00:
Because I have the same question as Judge Stoll.

[00:23:38] Speaker 00:
I feel like we've said repeatedly, if it's an arguable innovation that improves the functioning of the computer as a computer, have we ever said, well, if you get there,

[00:23:53] Speaker 00:
through an abstract idea or a non-technical idea or something that works for humans in other contexts, that's an exception to our point that if it improves the function of the computer, you're fine under Alice.

[00:24:06] Speaker 04:
Well, the court's holding are so narrow that as long as it improves technology or computer, that's sufficient to pass 101.

[00:24:14] Speaker 04:
That's definitely not the case.

[00:24:15] Speaker 04:
I tend to look at it as it just depends what perspective you're looking at.

[00:24:19] Speaker 04:
I feel like these 101 cases come before the court all the time.

[00:24:22] Speaker 04:
in a parallel processing, distributing, distributed networking.

[00:24:26] Speaker 04:
And the patent owner argues, well, it's a technological improvement.

[00:24:30] Speaker 04:
And then the court has held some of these decisions.

[00:24:34] Speaker 04:
Well, no, if you just arrange it in a way that mimics the abstract idea, that's not sufficient.

[00:24:39] Speaker 04:
Like the Eolis case versus Amazon, I think that's not presidential, but that was an example of one of the distributed processing cases that comes to mind.

[00:24:49] Speaker 04:
Accenture.

[00:24:51] Speaker 04:
that had like a task engine, I think a task database, all sorts of similar computer components.

[00:24:57] Speaker 00:
Do you think these are cases where we have said we grant that it improves the functioning of the computer, but because it does it based on an arguable abstract idea, not eligible?

[00:25:09] Speaker 04:
I don't know if the court came out and said, we grant that it improves the technology.

[00:25:13] Speaker 04:
But I don't think there's any dispute that there was sort of improvements in the processing.

[00:25:17] Speaker 04:
Like SAP is another example.

[00:25:20] Speaker 04:
And for me, it's just a matter of sort of what's the perspective that you look at it.

[00:25:26] Speaker 01:
Do you want to talk about your?

[00:25:28] Speaker 04:
Yes.

[00:25:29] Speaker 04:
So turning to the 275 patent, here the claims require co-processors that proactively retrieve a task from the task pool.

[00:25:38] Speaker 04:
Amazon argued in its petition that it would be invalid under any reasonable construction because we looked at the specific expressed examples in the patent specification.

[00:25:47] Speaker 04:
We figured the prior art does that and so no construction is needed.

[00:25:51] Speaker 04:
The board adopted a construction that neither party proposed.

[00:25:55] Speaker 03:
Didn't it say proactive means not reactive?

[00:25:59] Speaker 03:
I mean expressly, that's what it said.

[00:26:02] Speaker 04:
Correct.

[00:26:02] Speaker 04:
The board's construction is initiating change rather than reacting to events, i.e.

[00:26:07] Speaker 04:
not being told to act or not react.

[00:26:09] Speaker 03:
What's wrong with just setting that as everything else that you're arguing aside?

[00:26:14] Speaker 03:
Do you agree that it's reasonable to say proactive means not reactive?

[00:26:19] Speaker 04:
I think that's reasonable in the abstract, but if it's then, whether it's further interpreted or applied to exclude the express examples in the specification, which we'd say are also proactive, are not reactionary, then it's insufficient.

[00:26:33] Speaker 04:
So I think those plain words, though, are OK.

[00:26:39] Speaker 03:
OK.

[00:26:40] Speaker 03:
And when you say when it's applied, that's OK.

[00:26:44] Speaker 03:
concern I have is that I feel like your argument on claim destruction squishes over into the fact inquiry.

[00:26:51] Speaker 03:
You know, first to interpret the claims, then you're going to compare the claims as interpreted to the accused vice or the prior art, right?

[00:26:59] Speaker 03:
And so I feel like you've done that here.

[00:27:04] Speaker 03:
And that when by saying, first the court says or the board says proactive means not reactive, and then it applies it, why is that not a question of fact?

[00:27:15] Speaker 04:
Because if the specification says proactive includes A, B, and C, and the board adopts a construction that uses totally different words and says, oh, but that doesn't include example A, B, or C in the specification.

[00:27:29] Speaker 04:
We think that's an error of claim construction.

[00:27:30] Speaker 04:
We think whatever the construction is, any reasonable construction, as we said in the petition, well, it has to at least include these examples.

[00:27:37] Speaker 04:
Otherwise, that's not a reasonable construction.

[00:27:40] Speaker 04:
And that didn't come to really light until you see how they then apply those words.

[00:27:44] Speaker 04:
So we think that's why it is a legal question.

[00:27:48] Speaker 04:
If the construction excludes the specification examples, that's an improper construction.

[00:27:52] Speaker 03:
But isn't it the application that excludes those examples?

[00:27:57] Speaker 04:
I don't think so.

[00:27:58] Speaker 04:
I mean, here it's, I guess, tricky because the board is construing and applying.

[00:28:01] Speaker 04:
There's no jury here.

[00:28:03] Speaker 04:
But if the board, in adopting that construction, has decided that that construction excludes these examples,

[00:28:11] Speaker 04:
I think whether that's an error of claim construction or lack of substantial evidence, I don't know.

[00:28:15] Speaker 03:
It's hard when the parties don't ask for construction to know where construction begins and ends and what the party's positions are.

[00:28:24] Speaker 04:
We looked at the specification.

[00:28:25] Speaker 04:
We said, we do what's in the specification.

[00:28:27] Speaker 04:
Swarm didn't propose the construction adopted by the board.

[00:28:30] Speaker 04:
When you look at what they argued at the hearing at A9062, A9064,

[00:28:37] Speaker 04:
They say in each of those cases, it's sending an agent to retrieve a task based on its own internal determination.

[00:28:43] Speaker 04:
We said periodically, that was an informal explanation of the term.

[00:28:47] Speaker 04:
Their construction at A6488 and their response used one of those examples in the specification.

[00:28:54] Speaker 04:
I think all along, everyone was saying, well, if you look at the specific examples in the specification, that should be enough.

[00:29:02] Speaker 04:
That's all we're saying.

[00:29:04] Speaker 01:
Did you propose a form of words as a construction that made clear the inclusion of the examples?

[00:29:17] Speaker 04:
We didn't propose specific words.

[00:29:18] Speaker 04:
What we said is under any reasonable construction.

[00:29:20] Speaker 04:
And we specifically referred to this periodic going back example.

[00:29:25] Speaker 04:
That's at A4303 through 06 of our petition, our expert

[00:29:31] Speaker 04:
I'll say it one more time, so A4303-306.

[00:29:35] Speaker 03:
That's where your proposed construction is.

[00:29:37] Speaker 04:
That's in our petition and the Lowenthal Declaration at A5006 is one example where we argued that.

[00:29:50] Speaker 04:
But even if you move beyond the claim construction point to Judge Stoll just briefly say, we think if you look at, there's many tasks that are retrieved.

[00:30:00] Speaker 04:
All that happens is at the initial when a TCW registers, there's a notification the tasks are available.

[00:30:07] Speaker 04:
Analogous to in the Scrum board, you hire workers and you say, here's the board that has the tasks, that's it.

[00:30:12] Speaker 04:
And then thereafter, whether it's the first task or task two through a hundred, there's no additional notification that happens and that that should be

[00:30:21] Speaker 04:
sufficient because if the prior art validates it sometimes, or it would be obvious sometimes, then that would be sufficient.

[00:30:28] Speaker 00:
But what's unreasonable about the board's interpretation that because the whole thing only started with the kickoff, which with saying, in effect, here's the scrum board, I hire you, get started,

[00:30:41] Speaker 00:
I that nothing that follows is is proactive.

[00:30:46] Speaker 00:
It's all at some level a reaction to hey, I've hired you and here's the tasks get to work.

[00:30:52] Speaker 04:
I think that that reading doesn't match the examples of proactive in the specification.

[00:30:58] Speaker 04:
That would start to bleed into an overlap with how the specifier describes autonomous.

[00:31:03] Speaker 03:
Why would the examples of proactive in the specification, or the examples in the specification, why are they limit?

[00:31:11] Speaker 03:
Why should we read those into the claim?

[00:31:13] Speaker 04:
I'm not saying read them into the claim.

[00:31:15] Speaker 04:
What I'm saying is, here, and take the strum board example.

[00:31:20] Speaker 04:
You hire workers and you say, here's the board that has all the tasks.

[00:31:24] Speaker 04:
And then thereafter,

[00:31:26] Speaker 04:
What I would say, proactive orders, workers, are ones that periodically, in the words of the specification, go up and take tasks.

[00:31:31] Speaker 04:
They don't need to be told each time, hey, there's a task, hey, there's a task, you're not taking enough tasks.

[00:31:36] Speaker 04:
And so what we're saying is the construction excludes those specifications.

[00:31:41] Speaker 03:
Well, let's assume for a minute that we agree with the construction that proactive means not reactive.

[00:31:46] Speaker 03:
Okay, so what do you do?

[00:31:48] Speaker 03:
It's kind of like a thought experiment, if you will.

[00:31:52] Speaker 03:
You've got a circumstance where someone tells you to do something, and then you do it, and you do it again and again.

[00:32:01] Speaker 03:
Are you being proactive, or are you responding to what you were told to do?

[00:32:07] Speaker 03:
Being the initial kickoff.

[00:32:09] Speaker 04:
Well, I would say that example is different than Laurie.

[00:32:12] Speaker 04:
Laurie isn't saying, and there's nothing in Laurie that says,

[00:32:16] Speaker 04:
upon registration, the ATD instructs the TCW to go retrieve a task.

[00:32:21] Speaker 04:
So the example of you're saying if it's telling you to do something and then you do it, is that sufficient?

[00:32:25] Speaker 04:
That's not really what Laurie's doing.

[00:32:27] Speaker 04:
Laurie is just saying a notification to all TCWs, tasks are available.

[00:32:33] Speaker 04:
And then the question is thereafter, does that preclude all proactivity?

[00:32:37] Speaker 04:
We'd say no, as long as it's doing what the specification says.

[00:32:41] Speaker 04:
Periodically, going to get a task or

[00:32:45] Speaker 04:
The other example that Swarm relied on, when the TCWs free up and have processing, that's when it goes to get a test, and it's not waiting to be instructed or notified again.

[00:32:56] Speaker 04:
we'd say that is sufficient to be proactive.

[00:32:58] Speaker 03:
And you would have to say, if I say under the construction, the construction of proactive is not reactive, period, you would say that having an initial kickoff, no reasonable fact finder could say that the initial kickoff and then followed by the worker doing the task over and over and over again, that no reasonable fact finder could say that that's being reactive.

[00:33:26] Speaker 04:
Well, correct you under the substantial evidence, yeah.

[00:33:30] Speaker 04:
Yeah, we agree with that.

[00:33:30] Speaker 04:
And I mean, we think that's the case for the first retrieval, because it's not really an instruction.

[00:33:35] Speaker 04:
But even setting that aside, retrievals 2 through 100, we don't think that notification precludes all proactively for all time.

[00:33:42] Speaker 04:
And I think when you read it like that, it starts to bleed in what they argued autonomous means and what the specification says autonomous means.

[00:33:48] Speaker 04:
It kind of collapses those two terms.

[00:33:54] Speaker 01:
Thank you for your argument.

[00:34:01] Speaker 02:
Thank you, Your Honor.

[00:34:02] Speaker 02:
I want to address the 112 point first.

[00:34:05] Speaker 02:
And you're looking for support for the dynamic addition of the coprocessor.

[00:34:12] Speaker 02:
That is at appendix 155, column 10, at line 13 to 14, where it says, the wireless connection of cells into the system 10 facilitates dynamic addition and or removal of cells for the use of the system.

[00:34:31] Speaker 02:
And that also goes along with further language I previously quoted in column 10 that starts at 29 to 35.

[00:34:41] Speaker 02:
And then there's some more language about the wired configuration, which means that actually, let's look also at column 12.

[00:34:53] Speaker 02:
And in column 12, line six, I believe it says,

[00:34:58] Speaker 02:
It talks about the laptop connecting to the task pool, and then further down at lines 14 through 21, it talks about a nearby, unutilized device, such as a smartphone, connects to the task pool and sends its agents to fetch a masking task.

[00:35:16] Speaker 02:
Those are the dynamic support for the dynamic addition of additional co-processors.

[00:35:21] Speaker 02:
Also, Your Honor, I'd like to point out that I think my co-counsel conceded the obviousness position

[00:35:28] Speaker 02:
because he said that what Lurie does is the ATD notifies the TCWs that tasks are available.

[00:35:38] Speaker 02:
Our obviousness position is that claims 2 through 12 require searching the task pool.

[00:35:46] Speaker 02:
They require the co-processors to search the task pool.

[00:35:50] Speaker 02:
And here, Your Honors, nothing, nothing in Lurie searches the ATD.

[00:35:56] Speaker 02:
The ATD, what it does, what it says it does, it's an automated task distribution system.

[00:36:02] Speaker 02:
Thank you, Your Honors.

[00:36:03] Speaker 02:
If you have no more questions.

[00:36:04] Speaker 00:
Could I ask you on the cross appeal, on the proactive, do you concede that the board's construction reads out of the scope of the claims, at least the embodiment of Figure 6?

[00:36:17] Speaker 02:
No, Your Honor, we don't concede that.

[00:36:21] Speaker 02:
And we rest on our briefs on that matter.

[00:36:25] Speaker 02:
Thank you.

[00:36:28] Speaker 01:
There was no real argument on the cross appeal.

[00:36:33] Speaker 01:
Okay.

[00:36:34] Speaker 01:
Thank you.

[00:36:36] Speaker 01:
Case is submitted.

[00:36:37] Speaker 01:
Thanks to all counsel.