Part of the problem with all this is that AGI by definition is equivalent to writing a program that will "do everything," and that is why we get very general definitions of functionality like "solving complex problems with limited resources" and the like. The whole movement was to buck constraints. That doesn't mean there are no constraints, only that they are much more general. Mike A
On 4/10/13, Mike Tintner <[email protected]> wrote: > Ben: > If people were banned from doing science until they had a proof of > concept, how would they ever do the work needed to create the proof of > concept? ;-p > > Ben’s frivolous reply helps me frame a serious proposal – and IMO – an > extremely good idea for any AGI/inventor’s project AND FORUM! I hope > people will take it seriously because it could be enormously helpful to > creatives everywhere as well as AGI. > > Let me first recap: > > DEFINITION OF A PROPER INVENTIVE PROJECT > > Any seriously undertaken AGI or inventive project should have: > > 1) an **operational definition** – of the practical EFFECT(s) the machine > will produce – what it will do – in this case: what AGI problems it will > solve, (and how it will diversify to solve more AGI problems) > > 2)a **proof of concept** – which sets out an EFFECTIVE MECHANISM to produce > the desired effect - you must be able to give a practical reason why your > project will work – in this case your project’s mechanism to solve AGI > problems. It should be practically clear that that mechanism has a *chance* > of success/working – producing the desired effect (as a wine-press can > clearly press printed seals down). By definition, a machine involves many > mechanisms – the proposal will only cover one or two – and so may not solve > the others and work in the end. But it should constitute a definite start. > > By definition also, re Ben’s frivolous evasion, we are talking about what > should happen when inventors/creatives seriously **commit** to a project – > and especially when, like Ben, they involve others in their schemes. Before > then, they can think about any aspect of the project they like. But **if > they are going to get serious,** they should have both 1) and 2). > > Any serious creative person will *want* to impose such a definitional demand > on themself. The above requirements are common sense – a more specific way > of saying “you must define the problem (and outline at least a part of the > solution)” > > *WHY THE ABOVE DEFINITIONS ARE EXCITING* > > The non-philosopher/psychologist may not appreciate them, but these > definitions/terminology have the power to pin down both creative projects > and discussions, where existing definitions allow them to spiral into > confusion. > > For example, one common way of expressing 2) is to say “Ben doesn’t have an > *idea* of how to solve AGI *. This is true, but “idea” doesn’t pin down what > is missing, and can be endlessly evaded, arguing about what an idea is . > Ditto it’s common to say “Ben is offering a *magic sauce* for AGI “.. True, > but it still doesn’t really pin down what is missing, because it talks of > “sauce” as opposed to “effective mechanism”. > > Ask Ben : > > 1) what is Opencog’s operational definition of AGI – what effect/ AGI > problems will it solve? > > 2) what is its effective mechanism for solving those problems? > > .and you won’t get an answer. You will probably get a waffly evasion : “it > will solve a child’s intelligence test” or “be a child scientist” or > somesuch. But persist with “what practical effect will it solve – what AGI > problems will it solve?” and Ben & others will depart the field. They > haven’t got an answer, and unfortunately they aren’t even interested in > trying to produce one. > > WHY CREATIVES NEED TO BE PINNED DOWN. > > Ben classically demonstrates the necessity to pin would-be creatives down - > which is that so many will evade the problem. It is human nature, when > faced with a difficult creative problem, to do what Ben (AND ALL OF US) do – > turn to easier problems. Don’t deal with the problem at all, deal with > other nonproblematic mechanisms of your machine – not the effective, > “take-off” mechanism, but some problem of logic or processors. Or write a > paper on an entirely different philosophical problem. Or do your washing. We > are all endlessly guilty of this – Ben, being so productive, just does it > more visibly than most. > > But what also happens is that would-be creatives can spend *years*, > **DECADES**, an **ENTIRE LIFETIME** never dealing with the problem – never > even trying to produce an idea about their effective mechanism . > > This is actually a tragic waste of life – and almost a disease. We need to > stop it . And we can stop it. > > A PROPOSAL FOR A TRUE CREATIVE/AGI FORUM > > The above definitions make it simple to take action – and pin creatives > down. > > What we need is a v. simple **review system** on this forum – and every > inventor’s forum. > > A different version of what already exists on many forums – which is the > opportunity to click on “Vote up/approve” or “Vote down/disapprove” posts. > > General approval/disapproval or grades will have no effect. > > Instead I suggest we have two responses - “O.D./NO O.D. - Tick either > ... E.M. /NO E.M... Tick Either [for Operational definition – effect / > Effective Mechanism]. > > Obviously, readers should only respond when a post contains or refers to an > AGI proposal. > > Anyone who has been subject to a review system knows that this will have an > effect. I can cite my somewhat tangential experience > hosting/renting-out-rooms on airbnb. when you get specific criticisms, it > really stings – and you do something about them. Specific reviews are > powerful. > > If Ben were to receive a shower of “No O.D.”/”No E.M.” responses, you > would find that – whereas he has been able to brush off the isolated > criticisms of individuals – he would start noticing. And it might take > time, but a miracle would happen – he would actually start addressing the > problem of AGI – the E.M. – directly. He would want to get a tick or two. A > lifetime of evasion would be halted. > > Ditto Jim. The waffle would slowly start disappearing – he would start > sharpening up his act – might even introduce evidence. > > We all would. Criticisms – institutionally, publicly framed criticisms – > affect people, as we have seen throughout society in many forms in recent > decades. > > You’re programmers – can you find a way of introducing such a system here, > where everyone can automatically tick posts on these criteria? > > I kid you not – this simple system will “save lives” – save years of wasted > life – and generally sharpen up debate and standards. > > And if it works here, it will spread. Maybe there is some equivalent system > on another forum. As a philosopher, aware of the philosophical problems > involved in reaching appropriate definitions of creativity, I would doubt it > but it’s possible. > > In the meantime, I or others can frame a standard post of reply – with “NO > O.D. “etc – that people could send individually. It might mean a flood of > responses to a given post – but we could easily delete, without opening > them. > > Please do take this proposal seriously. Review systems have proven > effectiveness. They may seem cruel at first – they are actually extremely > productive and helpful. Don’t sink into inertia . Help others and help > yourself. > > > > > > > ------------------------------------------- > AGI > Archives: https://www.listbox.com/member/archive/303/=now > RSS Feed: https://www.listbox.com/member/archive/rss/303/11943661-d9279dae > Modify Your Subscription: > https://www.listbox.com/member/?&; > Powered by Listbox: http://www.listbox.com > ------------------------------------------- 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
