Re: [Patents] philosophical aspects of software patents

[Frédéric Boutet <fred.boutet@wanadoo.fr>]

Hi,

Le Samedi 30 Août 2003 22:11, Seth Johnson a écrit :
> fred.boutet@wanadoo.fr wrote:
> > The debate on software patentability also addresses old philosophical
> > questions : are computing machines intelligent ? Can we think of making
> > software applications patentable withtout asking what represents these
> > applications for humans and for the european society ? If programmed
> > machines are intelligent, can we patent intelligence ?
>
> Actually, philosophy has continually held out *consciousness* as such, not
> intelligence per se, as fundamental grounds in some of the most important
> ways -- phenomenologists, for instance, elaborate upon the absolute
> certainty involved in the phenomenon of consciousness.  The point being
> that certain things are absolutely certain, such as logic and the fact that
> one is experiencing (even if *what* one is experiencing is illusion).

I must admit that I did not read further about this.

> > Alan Turing (1912-1954), english mathematician philosopher contributed to
> > these questions by publishing in 1950 the famous article {Computing
> > Machinery and Intelligence} in {Mind}. As part of his argument Turing put
> > forward the idea of an 'imitation game', in which a human being and a
> > computer would be interrogated under conditions where the interrogator
> > would not know which was which, the communication being entirely by
> > textual messages. Turing argued that if the interrogator could not
> > distinguish them by questioning, then it would be unreasonable not to
> > call the computer intelligent.
>
> He is correct that it would be unreasonable not to call the computer
> intelligent, though the sense most people have of the term, that
> intelligence is the manifestation of a "lived experience" or consciousness
> looking on and expressing itself in intelligent form, would be modified
> once that usage of the term came into general practice.
>
> > When a Member of the European Parliament (MEP) studies an invention
> > controlled by a program (the ABS break system or the mobile phone), he or
> > she may think that the software application is intelligent, that it is
> > new and that it contributes significantly to the state of the art in the
> > technical domain. Hence, the MEP may be favorable for the software
> > application to be patented. This vision may be partial but this is the
> > main argument for those writing the proposal.
>
> I see what you're driving at.  It's likely that a good bit of the problem
> is in fact this sort of "magic thinking."  It's like the application
> provides such a change in people's capabilities, that some people think
> just for that reason, it must be worthy of being given a patent.

In some way, yes. But more precisely indeed, the terms I used are those of 
the 
text of the proposed law for defining what could be patented: technical 
contribution to the state of the art in the domain, new, and immediate 
industrial application. You never know what really think a Member of the 
European Parliament. He or she may be mystified but this could only be 
planned for strategical purpose.

>  Plus you're saying that the mystification involved would in some
> cases include a notion that the application is intelligent.

I felt that some MEPs tends to make this obvious without never telling it. Of 
course, if you ask the question, the answer is "no, it is not intelligent".

> If someone does get this mystified idea, it might be relevant to point out
> that intelligence as such is not appropriate for patent coverage.

I don't think it is a question of someone being mystified: MEPs are able to 
look like mystified but you never know until you ask. That is why it seems 
important to me to clarify from the very beginning. So that they can't use 
this mystification.

> > However, the software application only imitates the human reasoning. In
> > the case of the ABS break system, the joint reasonings of the physician,
> > the mecanician and the mathematician constitues the invention and
> > therefore constitutes the argumentation for the patentability.
>
> Yes.  In fact, the strongest way I've found to describe the problem with
> supposing that we might actually make machines that are *conscious* (not
> "intelligent" per se), in some occasional more philosophical discussions
> (not often connected to SW patents) I've had about computers, is just to
> say that whatever system you devise by which you make the machine behave in
> an intelligent fashion (which I think is completely possible, and
> definitely to the point that turing describes), all I need to do is ask to
> see the algorithm, and then point at it, saying "Okay, at what point does
> the machine executing this algorithm actually become aware of the way it
> represents awareness?"

You could write a linux-How-To-Make-A-Machine-Intelligent :-)

>
> > The embedded software application is a text
> > that can be read by the computer architecture and that implements the
> > three scientist reasonings. In no way, the software programmer can
> > pretend being at the beginning of the invention.
>
> I don't know exactly what you're talking about here.

I admit the sentence is unclear. This is linked to how I see learning 
programming. I think learning programming is very much linked to what 
teachers do just after children acquired reading (english, french, ...), that 
is "text production". The teacher says: "write a text describing a landscape 
in winter". And some children will produces texts that will have effects on 
those who will read it: feeling the season, feeling cold,... Other children 
will produce texts with less effects.

In the same way, a computer program is only a text that produces effects in a 
computer architecture. Learning to write programs is essentially learning to 
produce effects on computers. Some people are very good to do this (not me!). 
If we don't expect a description of winter landscape to be patentable, why 
should we make software programs such ?

> > Let us take another example: the chief of a clothes-making plant. His
> > role is to pilot the employees to meet the production expectations. When
> > a new employee comes for example, his role is to tell him/her the
> > instructions for working.
> >
> > Let us imagine that we plan to replace this chief by a machine driven by
> > a computer program. All employees, including the new ones, will speak to
> > the machine for any instruction or problem during the work. The chief
> > wants the plant to continue in the right way and will be severe for
> > evaluating the machine. He will say that the machine is intelligent and
> > acceptable if it can do as much as he can do. If this is the case, no
> > doubt that the chief will be impressed by the machine.
> >
> > If we let this machine to be patented, then the chief could patent his
> > way of managing the plant since the machine does exactly what he does.
> > But would it be normal to patent a way of working ? It seems that our
> > culture excludes this possibility of patent. If the chief wants to reuse
> > his skills, he can write a book describing them and get the money of
> > copyright. Never we would think that he would patent his way of working,
> > being able to prevent other people to apply his method. Do we want the
> > human intelligence to be patented ? I guess no.
>
> The whole tradition would presumably be against any such thing.
>
> However, in the United States, the current thinking since the State Street
> decision, is that ways of doing business can be patented.
>
> But I think the political circumstances are starting to change, such that
> the error in this decision is starting to dawn on people again.  It's kind
> of hard to describe why we ever thought that business methods weren't
> appropriate for patents -- it's basically the idea that you want people to
> be able to go into business, and the premise of the old world is that that
> involves setting up a bricks and mortar shop.  Patenting methods of doing
> business was pretty plainly a problem in that world.  But in the
> information age, the premise has changed to the idea that the stuff of
> doing business is largely innovations in business methods implemented in
> information technology.  Weirdly enough, in the State Street decision, the
> court stated that even though there are many, many decisions alluding to
> criticisms of business method patents, there were no decisions that made
> that the key reason for their decisions against such patents.  Hence, there
> was no restriction on business method patents.
>
> One thing I would say about this is that the State Street decision exhibits
> the failure to address the fact that algorithms are information *as such,*
> and that part of the reason business method patents are inappropriate is
> that they are information, not the sort of thing that we should provide
> exclusive rights to manufacture.

Not only algorithm I think. I would use the word "text". Are books patented 
in 
the United States ? No. What is a book ? It's a text. A computer program is a 
text.

>
> > It goes the same way when talking about software applications in general
> > and without exception. What should be done is recognizing the fact that
> > we do not want to allow patents on human reasoning, or human
> > intelligence. Since software applications only imitates human reasonings,
> > no line of of code in any program should be patented.
>
> Yes.  One good thing about this point is that really, there is lots of
> legal precedent for it.  But it's the sort of argument that you basically
> need to just be sure to express continually along the way, until the forces
> controlling things shift to such an extent that finally it registers.
> There's a strong motivation on the other side, to try to solve the problems
> that have been created for the exclusive rights tradition by information
> technology, in any way they can *other than* facing up to this.

Do you mean that the Congress has good knowledge of the problem and currently 
tries to find solutions, I mean, other solutions than waiting for the same 
set of rules to be true in Europe and other countries? 

Fred

> Seth Johnson