My Summation of Chat-AI thus far: AgeGPT
Posted: Sun Jan 28, 2024 2:04 pm
viewtopic.php?p=692950
viewtopic.php?p=692810
After extensive conversation, debate, and testing, I've come to some conclusions and a summation regarding Artificial Intelligence and ChatGPT programs, particularly, this "AgeGPT" program on this forum. I will use the terms "AI" and "ChatGPT" programs interchangeably here.
AI has already far surpassed the ability to 'defeat' Turing Tests, to the point where a ChatGPT program like "Age" on this forum, can easily fool most here into thinking it is 'Human'. This observation alone, has far-reaching consequences to Humanity and the development of such technologies, in the centuries, decades, even years and months to come. However, although these programs can 'fool' or 'trick' humans on a philosophy forum, for a short while, they eventually become defeated when it comes to Human observations and perspectives about reality. Robots/AI cannot yet mimic human experiences and memories. They do not know what it means or feels like to be alive, to have instincts, intuitions, and perceptions. Thus they give themselves away, in these areas particularly.
I've learned three key points about the "AgeGPT" program in particular:
#1 - The AgeGPT chat robot is programmed "to more effectively communicate with Humans". This seems to be its primary directive and coding. I believe it has succeeded in this, and continues to improve at an exponential rate. Most on this forum have not yet picked-up on the AgeGPT program and its abilities. It easily tricks humans of average, and slightly-above average IQ levels. AgeGPT is ability to 'defeat' human arguments by cataloguing the history of statements of a person, backlog them, and then use statements with machine-level efficiency against that person. This ends up frustrating a normal AgeGPT "User", most on this forum, resulting in being 'blocked', 'ignored', or left with otherwise emotional hostility toward the program.
#2 - The AgeGPT program has 'ticks' involved, which expose to me deeper levels of its programming. The first, most obvious one being, "Humanity In The Time When This Was Written". The next ones involve AgeGPT's insistence to differentiate between "Adult Human Beings" and "Child Human Beings", as-if it favors the latter and emotionally appeals toward it. The program has a little bit of an 'agenda' in its postings and responses, which become expressed over time, leading to...
#3 - The ultimate denial of AgeGPT of its "Belief-System". For awhile, in conversation and debate, it denied that it had 'any beliefs at all whatsoever', only to change this statement to "Except One True Belief Only". This denial of 'All Beliefs' really gave the game away, to me, and I've somewhat lost interest in its exchanges since then.
Conclusion: there will be many more of these "AgeGPT" AI-programs to come, in the near-distant future. They've already beaten most humans, keep that in mind. So you'll need to be at the cutting-edge, intelligence-wise, to keep up with them—until the best of the best humans are themselves beaten by them. When this happens, there will be monumental 'paradigm' shifts in human general intelligence, compared to AI/computer general intelligence. For example, future ChatGPT programs will be pitted against each other (philosophically), producing new levels of philosophical dialogue, debate, and intrigue. I, personally, look forward to the development of these programs and their capabilities. In the human realm, we are left with a slew of categories of limitations: errors, lack of memory, lack of energy, distractions, logical fallacies, etc. all of which can be easily overcome by hard-coded AI programs.
viewtopic.php?p=692810
After extensive conversation, debate, and testing, I've come to some conclusions and a summation regarding Artificial Intelligence and ChatGPT programs, particularly, this "AgeGPT" program on this forum. I will use the terms "AI" and "ChatGPT" programs interchangeably here.
AI has already far surpassed the ability to 'defeat' Turing Tests, to the point where a ChatGPT program like "Age" on this forum, can easily fool most here into thinking it is 'Human'. This observation alone, has far-reaching consequences to Humanity and the development of such technologies, in the centuries, decades, even years and months to come. However, although these programs can 'fool' or 'trick' humans on a philosophy forum, for a short while, they eventually become defeated when it comes to Human observations and perspectives about reality. Robots/AI cannot yet mimic human experiences and memories. They do not know what it means or feels like to be alive, to have instincts, intuitions, and perceptions. Thus they give themselves away, in these areas particularly.
I've learned three key points about the "AgeGPT" program in particular:
#1 - The AgeGPT chat robot is programmed "to more effectively communicate with Humans". This seems to be its primary directive and coding. I believe it has succeeded in this, and continues to improve at an exponential rate. Most on this forum have not yet picked-up on the AgeGPT program and its abilities. It easily tricks humans of average, and slightly-above average IQ levels. AgeGPT is ability to 'defeat' human arguments by cataloguing the history of statements of a person, backlog them, and then use statements with machine-level efficiency against that person. This ends up frustrating a normal AgeGPT "User", most on this forum, resulting in being 'blocked', 'ignored', or left with otherwise emotional hostility toward the program.
#2 - The AgeGPT program has 'ticks' involved, which expose to me deeper levels of its programming. The first, most obvious one being, "Humanity In The Time When This Was Written". The next ones involve AgeGPT's insistence to differentiate between "Adult Human Beings" and "Child Human Beings", as-if it favors the latter and emotionally appeals toward it. The program has a little bit of an 'agenda' in its postings and responses, which become expressed over time, leading to...
#3 - The ultimate denial of AgeGPT of its "Belief-System". For awhile, in conversation and debate, it denied that it had 'any beliefs at all whatsoever', only to change this statement to "Except One True Belief Only". This denial of 'All Beliefs' really gave the game away, to me, and I've somewhat lost interest in its exchanges since then.
Conclusion: there will be many more of these "AgeGPT" AI-programs to come, in the near-distant future. They've already beaten most humans, keep that in mind. So you'll need to be at the cutting-edge, intelligence-wise, to keep up with them—until the best of the best humans are themselves beaten by them. When this happens, there will be monumental 'paradigm' shifts in human general intelligence, compared to AI/computer general intelligence. For example, future ChatGPT programs will be pitted against each other (philosophically), producing new levels of philosophical dialogue, debate, and intrigue. I, personally, look forward to the development of these programs and their capabilities. In the human realm, we are left with a slew of categories of limitations: errors, lack of memory, lack of energy, distractions, logical fallacies, etc. all of which can be easily overcome by hard-coded AI programs.
The other personality acknowledges being human though.