Matt, On Sun, Mar 24, 2013 at 4:36 PM, Matt Mahoney <[email protected]>wrote:
> On Sun, Mar 24, 2013 at 3:36 AM, Steve Richfield > <[email protected]> wrote: > > >> Do you really think that these companies don't already own hundreds or > >> thousands of patents on the methods they invented and are using? I > >> suppose if your price is low, they will pay it to avoid the hassle. If > >> you ask for too much, they will point out prior art to invalidate your > >> patent. > > > > This sounds like a game I can't lose!!! If the patent won't stand > anyway, challenging them would get THEM to do the hard work of finding any > prior art, before I spend any more in this direction. If they fail, then I > will have vetted my approach and it WILL be worth lots of money. Hence, I > should demand a fortune, and let the chips fall where they may. The odds > would almost certainly be better than a lottery ticket. > > > > Do you see any flaws in this obviously non-conservative logic? > > Yes. It will be very expensive to file a lawsuit if you lose. > My plan didn't include my filing a lawsuit. If things were to turn in that direction, I would partner with someone else who has deep pockets and an interest in the technology, e.g. a competitor. > > Did you really look at all 8 million patents (including expired ones) > in the U.S. to see if your invention has prior art, It sounds like you haven't searched the USPTO.gov web search before. The standard search looks at ALL present and expired patents, back to the time they first started digitizing patents, as I recall, back in the 1970s. Also Google now also looks at ALL digitized patents. I like Google's search engine MUCH more than the USPTO's, so I now do much of my patent searching using Google. Just include the word "patent" in your search string. or even the 1 > million or so that relate to computer technology? What about foreign > patents? In software, that is pretty much zero, because the foreign restrictions on software patents are crushing. > And remember, that something doesn't even need to be patented > to be considered prior art. The patent office has published EXACTLY what can be considered to be prior art, e.g. if it was used in a product but its operation was hidden, it is NOT prior art!!! > It just has to be disclosed, as in a > research paper, technical document, or open source code. I would have to go back and read carefully, but I suspect that open source code would NOT be considered prior art, unless: 1. You could PROVE that it was around before the patent application, and 2. Its operation was explained OUTSIDE of the code in some publicly accessible way, e.g. on the Internet. Are you > absolutely, positively sure that you are the first to describe these > ideas? > Of course not. These days, no one can make such a claim. > > > The BIG question is whether it works, and whether others believe it. The > thing missing in my (and other) parsing approach(es) is a canonical form to > represent syntax and semantics, with plenty of hooks to attach new code to > do new things. BNF would require a **LOT** of extension. My present plan is > to get people from various areas to work on such a representation. I am now > creating a paper for WORLDCOMP about the next steps, the initial submitted > draft of which I will post here in a week or two for comment. > > WORLDCOMP will publish any paper as long as you pay the conference fees. > They have been accepting ~50%. > > >> http://en.wikipedia.org/wiki/JPEG#Patent_issues > >> The patent claim in this case was the "invention" of using a single > >> code to represent a run of zeros followed by a non-zero value. Who > >> would have thought it would be worth $105 million? The claim wasn't > >> even valid, due to prior art. > > > > They screwed up. The trick is to just sue medium-sized companies for > lunch money, until the patent lapses. THEN sue the big guys for past > infringement. Also, look for settlements to avoid testing the patent. > Further, make a deal with one of them to fund the suits, in return for a > free ride. That way, you collect the money that can be collected now > without a court hassle, and risk invalidation only after the patent lapses > and you have already collected all that you can collect without a court > hassle. > > You think Forgent could have made more than $105 million before the > USPTO invalidated the key claims of the patent, 6 months before it was > to expire anyway? > The last 6 months probably didn't make all THAT much difference, but why not wait until it has expired to sue? At least that way, you can collect from everyone who does NOT insist on being taken to court to be made to pay. > > >> You might want to pay attention to what Kurzweil is doing at Google. > >> He has put together a team of several top researchers to tackle > >> exactly the natural language problem. He has access to a model with > >> 300 million concepts, and an awful lot of computing power. > > > > Sounds like he is doing SOMETHING the hard way, which is a standard > problem with people who think they have SO much money that they don't have > to worry about scaling issues. Our hundred billion neurons is the result of > a hundred million of years of optimization. People who think they can > "throw iron" at AI problems are just wasting everyone's time. > > Google doesn't want to build a language model equivalent to a single > human. That would not be very useful. They want to build a model of > all of the billions of people that use the internet. > I should probably send them a letter and see if there is any connection with what I am doing. > > >> > Cyc will never ever do anything useful. > >> > >> I agree. So why are you proposing a rule based system too? > > > > The trick is to use rules to identify the exceptions and problems, and > ignore the rest of the real world that bogs down Cyc. That seems to be what > we do in our own brains, as even 100 billion neurons couldn't begin to be > able to track all the things that are working just as they should. > > Cyc is not bogged down by rule evaluation. Rules are only evaluated as > needed using forward and backward chaining. This worked quite > efficiently, even on 1980's computers. What bogs down Cyc is the > manual process of entering millions of rules. It's not just typing > them in, either. You have to make sure the rules are consistent, and > have a way of debugging them when you discover they aren't because > different people have different beliefs. What are the chances of > entering a million rules with no errors? > Yes, that is my understanding. I just didn't see the cost/benefit of the Cyc approach. At least with your approach, you end up working with the pieces that are part of the world's engineering and commerce, and NOT with all of the tiny little itsy bitsy details therein. Your approach seems much better than Cyc's for understanding the pieces of the world, rather than understanding just the errant, malfunctioning, or otherwise noteworthy pieces as my approach seeks to do. The ~7 orders of magnitude difference in estimated costs between the difficulty of my approach and your approach, a hundred million compared to a quadrillion, feels about right. As with many things, my approach is one seeking less with less. Steve ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657 Powered by Listbox: http://www.listbox.com
