Although AI cognitive faculty has gained improvement yet the AI ,currently,can not independently create real innovation without human input.

An AI system could search for gaps in art or in technology, and try to fill them by pursuing own goals. In a first step, such gaps must be identified, which is a pattern recognition task. For example, there is hunger in the world, and it is worth fixing this problem. The AI system could ask itself: "How to eradicate hunger by using X?", where X is a randomly chosen or otherwise identified tool. I put "How to eradicate hunger with sand?" into ChatGTP, and got a five point suggestion starting with

"Eradicating hunger with sand is an intriguing and innovative concept, often associated with modern agricultural techniques. The most notable method is the use of a type of treated sand known as "hydrophobic sand" or "nano-coated sand" to improve water retention in arid regions. Here's how this approach can contribute to eradicating hunger:"

Regards,

Joachim

Dariusz Prokopowicz

I do not know answers but I do have additional questions.

Traditionally, according to Mode 1 science, scientific truth (fact, hypothesis, rule, theory, etc.) was considered a "properly proved correct belief". According to socially-orientated and socially-extended Mode 2 science, which is more about human-created problematics than about naturally-existing phenomena, it has been reinterpreted so as a "collectively agreed useful belief" (which, as knowledge and/or plan allows understanding and resolving a complicated problematics at hand). Probably, artificial intelligence-triggered Mode 3 science will further deform the original and widely-accepted paradigm of scientific knowledge so as "large associations-enabled probable conjectures". I have no doubts about accepting as a fact that by covering a large part of the human-recorded intellect (signals, data, information, knowledge, meta-knowledge) and by being intensively supported by computing capacities incomparably larger than the sensing/retrieval/inferring/associative capabilities of human collectives, artificial intelligence approaches will be able to generate novel beliefs or conjectures. However, I (still) have heavy issues with understanding the truth aspect(s) of AI findings. First of all, should they be true? If so, according to what? Do we have a mental framework for this? Second, what will "properly proven" or "collectively agreed" mean in this case (because just being "large associations-enabled" does not bring these closer to any human assumpted forms of truth)? Third, while human semantics is involved in the verification/validation/consolidation of something as a truth, how "the trueness" will be discovered and accepted? By AI or by humans? If the first applies, then we are in a circulus vitiosus (i.e. in a fallacy involving the presupposition of a proposition in the premis ...). How do you think about my five cents?

Thanks for the facinating question. Kind regards,

I.H.

Dear Dariusz Prokopowicz ,

I appreciate that you have taken your time to give your views regarding opportunities that AI might bring about in regards to new inversions. It is useful to have your outlook on the subject, as well as your elaborate analysis of the use of generative AI and other sophisticated technologies.

Many peoples have differed in their opinion on AI’s ability to actualize distinct innovations on its own, given that there has been notable progress in its ability to work through datasets and create content. The heart of the issue is usually where the boundary between ‘new to the world’ and whether a product can be labelled new because it uniquely solves a problem in a specific area or business is drawn. Some critics still focus on the fact that an AI system works with data created by humans, which implies that its functions are limited to the processing of new knowledge based on the existing data containing applicable and tested patterns. Therefore, from my point of view, the ability to come up with groundbreaking ideas may still be limited to humans compared to AI and robots.

I would like to thank you for starting this discussion as you pointed out, it is crucial.

Best regards,

Yahya Layth Khaleel

Yahya Layth Khaleel

I share your view in that sense that probably much more is needed for a true 'eureka' (that is, for spontaneous independent context-driven inception, emergence, popping up, etc. of brand new notions, concepts, solutions, and/or designs that are valuable and feasible from a human perspective) than the procedural or algorithmic association of massive data, deriving and matching patterns, and generating vague textual circumscriptions. While, with all known inherent limitations, humans discover or invent, they simultaneously and intuitively also think about the meaning, feasibility, utility, trade-offs, norms, etc. of the discovered or invented something, even if their tested knowledge is incomplete and insufficient about it. This disposition is still missing from AI approaches, as far as I know. Ultimately, it leads us back to the hot (burning) topics of reproduced consciousness, the working of (intrinsic, ineffable, apprehensible, and private) qualia, and the physical basis of feel and sensation. My feeling is that, since we do not know enough about these last-mentioned things, we cannot make defendable statements about their role in artificial intelligence as a whole. I think this definitely applies in the context of discovering, inventing, or generating reasonable and useful unknowns to the human world, but probably also to any non-trivial incremental novelties. We must also keep in mind that, furthermore, there is also a teleological question hiding behind.

With best regards,

I.H.

Quite facinatinng line of thought but i believe that given the current state of the art, Ai would not be able to carry out any innovations on its own. That is where the human element is needed because AI needs humans to think and to produce.

I would like to give my opinion out of the scope given. Firstly, when u ask various AI u will never get the similar answer, if I'm right. In we ask what id the origination of Covid -19, the answer may be... to the best of my knowledge........ Various things in the world r not yet known actually, Pyramid is making for what n the reply is coming from nearly the similar first people who find it and deep assessment with the current circumstance, but not at the time of occurred. Meanwhile, others is Blackhole, gravity force still do not known with the current scientific data, the current conclusion is based upon the current factors but not from the time of origin even though it is the studying back based on the scientific data n valuable information from evidences and generate simulation conclusion. But certainly this is the best limited we can do right now.

In my opinion, the development is from the mutation only which I think Gen AI can do. At this time, the reply of each answer may be vary from AI to AI, but after times, AI will learn it all until all of them will get the same answer that will be no speculation any more or other things and may bring it back to everything decision is the same, everyone is similar with zero differs.

This is what I want to share from my studying, AI is the 4th Generation and the final is the 5th Generation indifferent in everything and people a given above.

Best regards,

JB.

Dear Dr. Dariusz Prokopowicz :

For days, I have been thinking about your second (i.e., the first explanatory) question: "Can artificial intelligence create innovations with the help of artificial intelligence, since the knowledge bases of AI applications contain what humans have already created before?" The reason is the easily observable duplication in the first part of the question, that is, using the term 'artificial intelligence' twice in a specific (in fact, implicit) semantic relationship - (which is, "with the help"). Is it just a ("tautological") wording issue or is there some meaning hiding behind that should be made explicit? The second part is clear. Please kindly provide information. Best regards,

I.H.