@olugt.bsky.social
Also, even though the overall guesswork impression of AI can especially be predictive in nature in its workings, it still does not transfer accountability to the AI because its design as taking predictive steps in its own perspective is a guesswork as overall perspective and impression.
June 17, 2025 at 6:31 AM
Also, even though the overall guesswork impression of AI can especially be predictive in nature in its workings, it still does not transfer accountability to the AI because its design as taking predictive steps in its own perspective is a guesswork as overall perspective and impression.
...don't take that someone seriously but keep the relating at acquaintances level, because you do not have reasons to inform a decision to enter into a serious relationship with that someone, much less marriage.
June 14, 2025 at 2:36 PM
...don't take that someone seriously but keep the relating at acquaintances level, because you do not have reasons to inform a decision to enter into a serious relationship with that someone, much less marriage.
In other words, if you do not have first-hand experiences to observe for good attributes in someone and in your relating with that someone, especially a whole combination of growth, responsibility and accountability attributes,...
June 14, 2025 at 2:36 PM
In other words, if you do not have first-hand experiences to observe for good attributes in someone and in your relating with that someone, especially a whole combination of growth, responsibility and accountability attributes,...
A dating app should not be going this far with browser history or whatever, because it's pointless. It's nonsensical.
June 14, 2025 at 2:31 PM
A dating app should not be going this far with browser history or whatever, because it's pointless. It's nonsensical.
To get into something serious relationship (much less marriage) a person would have to steadily encounter people and filter those people by good attributes and doing so using especially first-hand experiences.
June 14, 2025 at 2:31 PM
To get into something serious relationship (much less marriage) a person would have to steadily encounter people and filter those people by good attributes and doing so using especially first-hand experiences.
People don't realize that dating app should simply be used to encounter people, not something to use to expect to get a serious relationship or marriage out-of-the-box.
June 14, 2025 at 2:30 PM
People don't realize that dating app should simply be used to encounter people, not something to use to expect to get a serious relationship or marriage out-of-the-box.
However, AI makers cannot use the guesswork nature of AIs as an excuse to avoid accountability because an approach of accountability can be by implementing constraints, limits, etc. about the guessing that the AI does.
June 2, 2025 at 1:23 AM
However, AI makers cannot use the guesswork nature of AIs as an excuse to avoid accountability because an approach of accountability can be by implementing constraints, limits, etc. about the guessing that the AI does.
Whatever an AI does in general can be well understood, but what an AI does in particular or in specific scenarios (i.e. the guesswork) cannot be typically understood beforehand because guesses cannot be typically understood beforehand. That is typically the nature of guessing.
June 2, 2025 at 1:18 AM
Whatever an AI does in general can be well understood, but what an AI does in particular or in specific scenarios (i.e. the guesswork) cannot be typically understood beforehand because guesses cannot be typically understood beforehand. That is typically the nature of guessing.
Hence, proponents of AI and AI makers should stop shifting accountability to AI. AI makers should be accountable for what the AI does, instead of saying they do not understand what the AI does.
June 2, 2025 at 1:13 AM
Hence, proponents of AI and AI makers should stop shifting accountability to AI. AI makers should be accountable for what the AI does, instead of saying they do not understand what the AI does.
So, just as it is not typical to predict guesses, it is not typical to understand what an AI will guess unless initial conditions are known about each step of the guesswork the AI is doing. A computer can probably log the initial conditions for review, but it will be too tedious to manually review.
June 2, 2025 at 1:08 AM
So, just as it is not typical to predict guesses, it is not typical to understand what an AI will guess unless initial conditions are known about each step of the guesswork the AI is doing. A computer can probably log the initial conditions for review, but it will be too tedious to manually review.
Accountability of AI makers should include to ensure their AI does not guess into the territory of harmful or expensive mistake stuff, and this can be by implementing constraints, limits, etc. about the guesswork the AI is doing.
June 2, 2025 at 1:05 AM
Accountability of AI makers should include to ensure their AI does not guess into the territory of harmful or expensive mistake stuff, and this can be by implementing constraints, limits, etc. about the guesswork the AI is doing.
But with AI, it is guessing for what is most likely to be the correct thing. It is going to produce a different result each time.
This is an example of guessing systems.
This is an example of guessing systems.
June 2, 2025 at 12:57 AM
But with AI, it is guessing for what is most likely to be the correct thing. It is going to produce a different result each time.
This is an example of guessing systems.
This is an example of guessing systems.
It is, of course, complex but still concretely certain at the more macroscopic levels (e.g. beyond quantum level about how electrons work).
June 2, 2025 at 12:56 AM
It is, of course, complex but still concretely certain at the more macroscopic levels (e.g. beyond quantum level about how electrons work).
Everyone does not have to really understand how electrons or whatever move around to convert taping on their probably touchscreen keyboard to make a text that another person reads on their end on the Internet and that is same as the initial text.
That is an example of concrete systems.
That is an example of concrete systems.
June 2, 2025 at 12:52 AM
Everyone does not have to really understand how electrons or whatever move around to convert taping on their probably touchscreen keyboard to make a text that another person reads on their end on the Internet and that is same as the initial text.
That is an example of concrete systems.
That is an example of concrete systems.
I should clarify that of course there are useful "AIs" in research and other niches where I think they shouldn't have been called AI anyway, because there are more specific terminologies for such various advanced solutions/technologies. But my point is about the ones prioritizing negative impact.
May 25, 2025 at 3:35 PM
I should clarify that of course there are useful "AIs" in research and other niches where I think they shouldn't have been called AI anyway, because there are more specific terminologies for such various advanced solutions/technologies. But my point is about the ones prioritizing negative impact.
Hence, leading to very much less unprecedented losses of jobs and businesses. The economy is for and because of humans. Not AI.
May 24, 2025 at 10:22 PM
Hence, leading to very much less unprecedented losses of jobs and businesses. The economy is for and because of humans. Not AI.
Another is heavy relevant regulations on AI. The extreme harms that AI can cause or causes are basically why there are regulations and systems even for existing wrong or anticompetitive practices prior to the use of anything AI, much less the use of AI on the scale it's been in recent times.
May 24, 2025 at 10:20 PM
Another is heavy relevant regulations on AI. The extreme harms that AI can cause or causes are basically why there are regulations and systems even for existing wrong or anticompetitive practices prior to the use of anything AI, much less the use of AI on the scale it's been in recent times.
Starting point should be accountability. Holding AI makers accountable for making tools that inherently allow and aid wrongdoing in unprecedented and likely untraceable ways.
This is clearly different from when a knife made for the kitchen is used by a wrongdoer. The crime can be traced enough.
This is clearly different from when a knife made for the kitchen is used by a wrongdoer. The crime can be traced enough.
May 24, 2025 at 10:11 PM
Starting point should be accountability. Holding AI makers accountable for making tools that inherently allow and aid wrongdoing in unprecedented and likely untraceable ways.
This is clearly different from when a knife made for the kitchen is used by a wrongdoer. The crime can be traced enough.
This is clearly different from when a knife made for the kitchen is used by a wrongdoer. The crime can be traced enough.
The free accounts are a huge part of the issue. Free things are dangerous because the motive of the giver can be wrong now or later. And today's events are typical examples. If everyone paid, the platforms would not have had this much data they can twist legalities about for selfish purposes.
May 24, 2025 at 9:49 PM
The free accounts are a huge part of the issue. Free things are dangerous because the motive of the giver can be wrong now or later. And today's events are typical examples. If everyone paid, the platforms would not have had this much data they can twist legalities about for selfish purposes.
In contrast, in serious programming, software engineering, etc., things are usually done relatively from scratch to ensure things turn out as exactly as intended as possible. Hence, even the tooling platforms being used stick to being tools to interpret what the programmer intends. Lots at stake.
May 24, 2025 at 9:46 PM
In contrast, in serious programming, software engineering, etc., things are usually done relatively from scratch to ensure things turn out as exactly as intended as possible. Hence, even the tooling platforms being used stick to being tools to interpret what the programmer intends. Lots at stake.
Of course, some limitation are there, like of AI not being able to reproduce exactly what the human creative intends. But albeit good enough for nonchalant people.
May 24, 2025 at 9:46 PM
Of course, some limitation are there, like of AI not being able to reproduce exactly what the human creative intends. But albeit good enough for nonchalant people.
A main reason why AI can negatively impact jobs in the creative and media sector is because their use of software tools are creatively clicking and designing things albeit with tooling already available in the various software tools. A lot of this can be abstracted away, which some AIs are doing.
May 24, 2025 at 9:45 PM
A main reason why AI can negatively impact jobs in the creative and media sector is because their use of software tools are creatively clicking and designing things albeit with tooling already available in the various software tools. A lot of this can be abstracted away, which some AIs are doing.
In contrast, in serious programming, software engineering, etc., things are usually done relatively from scratch to ensure things turn out as exactly as intended as possible. Hence, even the tooling platforms being used stick to being tools to interpret what the programmer intends. Lots at stake.
May 24, 2025 at 9:40 PM
In contrast, in serious programming, software engineering, etc., things are usually done relatively from scratch to ensure things turn out as exactly as intended as possible. Hence, even the tooling platforms being used stick to being tools to interpret what the programmer intends. Lots at stake.
Of course, some limitations are there, like of AI not being able to reproduce exactly what the human creative intends. But albeit good enough for nonchalant people.
May 24, 2025 at 9:35 PM
Of course, some limitations are there, like of AI not being able to reproduce exactly what the human creative intends. But albeit good enough for nonchalant people.