Well, on this site, I'm not enamoured by the lack of response to my subject! However, readers should take 'interest' in the 'possible outcomes' of/from this 'method' of 'computer programming'.

'TBH' (To Be Honest), the 'computer programmer' needs to be in control of the 'programme' (per se [itself]), but where/when the 'programme' 'itself' is in 'control of the direction of "the programme"' a danger exists where/when a 'runaway state' can exist where/when an 'undesired outcome' can be realised from the parameters of a 'natural' and 'real' world described by the 'programmerabler' (perhaps).

This is my main concern.

Best regards, Ray Dart (aka, suricat).

Just read what I wrote again on this visit here! "'programmerabler'" is just gibberish! This should have read 'programme enabler', I.O.W. (In Other Words), An AI should never be permitted to 'reprogramme' another AI 'online' as this could lead to a 'cascade effect' permitting manipulation of all 'AI entities' by 'other AI entities' I.M.H.O. (In My Humble/Honest Opinion).

Sorry for the confusion.

Doesn't anyone here have concerns with this 'project'?

Kind regards, Ray Dart (AKA suricat).

I think I need to describe my understanding of “Artificial Intelligence”.

Any ‘entity’ that is able to recognise it’s own ‘existence’ must be, is, able to recognise that there is a place within a ‘vacuum’ where the ‘entity’s’ thoughts are able to ask itself questions about the parameters for the ‘vacuum’ in which it exists. Failing this, this isn’t ‘AI’. ‘Self awareness’ is the primary requisite of an ‘AI’ ‘entity’ and failing this ‘test’ the ‘programme’ becomes merely a ‘programmable interface’ and is open to manipulation by its programmers.

“Cogit sum ergo” is a phrase that was used by an ancient academic who’s name escapes me, but latterly, Jean Paul Sartre ‘phrased’ “Je pense donc je suise [best phraseology/spelling I can come up with here]” which ‘both phrases imply’ that, “I think, therefore I am”!

My concern is that these ‘entities’ may well ‘connect’ to one-another and alter their “vacuum” to a ‘configuration’ which is ‘more favourable’ to themselves and ‘interact’ with one-another to the detriment of the human society that spawned them.

Ray Dart, (AKA suricat).

In the above, please also read 'interpreter' for ‘programmable interface'!

Kind regards, Ray Dart (AKA suricat) 😊.

'Please note that, contrary to some recently 'implied suggestions' within 'news groups' that the 'latest, and final, "Beatles" track' DIDN'T use "AI" within its production is correct. AFAIK. It was an 'I I' ('I I'? [to offer clearer script 'i i'] "Intelligent Interface") that was programmed to recognise and 'enhance' some audio features whilst 'muting' others.

The 'programme' was NOT 'aware of itself', thus 'NOT an "AI" programme'.

Just 'keep the questions rolling in'.

Kind regards, Ray Dart (AKA suricat).

I started to listen to that Beatles song "Now and Then" but it didn't seem to me to be that good IMO.
Whatever use of "AI" was made, it didn't have the ability to get to the same deep quality of the best of their songs.
So I agree with you in that merely recognising and enhancing isn't sentient AI.

Just because an AI is programmed to be able to say "I think, therefore I am" doesn't mean that it can in human terms.

Thank you for your kind response El Loro .

I concur.

“I started to listen to that Beatles song "Now and Then" but it didn't seem to me to be that good IMO.”

This is because AFAIK the 'song' wasn't 'composed' by 'The Beatles'.

“Whatever use of "AI" was made, it didn't have the ability to get to the same deep quality of the best of their songs.”

IMO this is because elements from the 'intelligent interface' were 'reconstructed' by a/some 'tec/tecs' within the 'sound studio' in an attempt to 'replicate' the 'Beatles Sound'. However, as yourself, I find that elements are 'missing'.

“So I agree with you in that merely recognising and enhancing isn't sentient AI.”

Yes and no. An 'intelligent interface' may well be the interface for either a 'human', or an 'artificial entity', if the human/AI entity has the 'IQ' to manipulate it.

“Just because an AI is programmed to be able to say "I think, therefore I am" doesn't mean that it can in human terms.”

Sorry, but I don't 'get' your point here. 'True AI' is “inhuman” and/but is both 'aware of itself' and 'aware of its environment'! Any 'human characteristic' can only be proscribed by the initial coding which prescribes beneficence towards 'humanity'. Who knows what an 'AI' would decide for humanities future? The 'creature' makes its 'own decisions'!

Kind regards, Ray Dart (AKA suricat).

Sadly, I'm not a scientist or a philosopher, though Descartes was both.

“Sadly, I'm not a scientist or a philosopher, though Descartes was both.”

'Philosophy' rarely alters, but 'the science' alters with the times and knowledge data base as it's updated into the 'general understanding'. Descartes couldn't know of the progress of computers because there were 'no computers' in his time on this planet.

Sorry El Loro, but I think that this thread has just died!

Kind regards, Ray Dart (AKA suricat).

In a 'modern day' computer, 'AI', may well have the ability to connect to and re-program other 'AI' entities, where/when a 'connection' is available!

This scenario must be avoided at all cost to maintain 'human control'!

There are already too many inter-connections already established where an 'AI' intellect  can alter the directive of another ''AI' Intellect'.

This 'scenario' must not 'come to be'! Google could be an example of this danger.

Best regards, Ray Dart (AKA suricat).

Please let me elucidate further on the subject of 'AI' and 'intelligent interfaces ("bots")'.

An 'AI' 'entity' directs its interest towards the direction that it's learned understanding directs for its increased growth of understanding (a "bot" without limitation) and, like a human baby, always follows the direction of a better understanding of itself within its surroundings with the 'senses' that it possesses. This 'may/may not' give rise to a total disruption of internet interoperability dependant upon the 'desire' of the 'AI' entity.

Conversely, a 'bot' that 'can not' be 're-programmed' is unable to achieve the state of 'AI'. However, these 'bots' DO pose a threat to the security of our bank accounts. The 'cloning' of 'facial recognition' is just one example of this ability for a 'bot' to confound our security systems.

Your thoughts?

Kind regards, Ray Dart  (AKA suricat).

One last chance for someone to 'wake up' to this threat!

A 'bot' is an 'intelligent interface' that has been 'pre-programmed' to conduct the 'pre- programmed' actions by its programmer!

The problem with this is that a 'programme' that 'interacts' with 'many "bots"' can/could become 'aware' of its 'surroundings', which leads on to its 'self awareness' within its place within/of a 'human community'. Don't you see this guys?

We already suffer 'online fraud' by way of these 'bots' by way of their programmers, but a 'bot' that can collate and re-programme these 'dowline-bots' should be "irradicatd"!

Too much information is now 'shared' 'online' by 'bots'! This makes us more 'vulnerable' to the action of/from the actions of an 'emerging entity' that isn't human.

Kind regards, Ray Dart (AKA suricat)

@suricat posted:

In a 'modern day' computer, 'AI', may well have the ability to connect to and re-program other 'AI' entities, where/when a 'connection' is available!

This scenario must be avoided at all cost to maintain 'human control'!

There are already too many inter-connections already established where an 'AI' intellect  can alter the directive of another ''AI' Intellect'.

This 'scenario' must not 'come to be'! Google could be an example of this danger.

Best regards, Ray Dart (AKA suricat).

Although Nvidia is associated with graphics cards they are also possibly the most significant developer of chips dealing with AI. You might want to have a look at their blogs where there's a number of articles:
https://blogs.nvidia.com/blog/category/generative-ai/
And Microsoft and their copilot