38 replies to this topic

[m468]

Something designed to think logically will have a hard comming up with an illogical solution to a problem that may not be solved other wise

[Zero]

Well Smart AIs are made from humans brains so technically it's a person trapped inside a computer.

[KhevaKins]

Well Smart AIs are made from humans brains so technically it's a person trapped inside a computer.

More like a copy of a dead human brain with most of the memories stripped away but retaining some experiences.

[D4RKST0RM99]

Just occurred to me how differently the Keyes loop would have turned out had the Iroquois's AI been installed Keyes may have relied on its input and possibly would never have considered the insane manoeuvre he came up with as more than likely it would have advised against it and suggested a more logical and less risky approach if any.

I would not leave an AI in the same command position of a human captain, to give a better explanation to this reasoning if anyone has seen the Will Smith- I Robot movie a flashback scene reveals his character's backstory where he was involved in a car crash and his car is sinking while pinning a second car containing a trapped young girl, both would have perished were it not for a passer-by robot that witnessed the crash and attempted to save them, there was only enough time to save one of them based on their probability of survival and the robot chose Will Smith's character leaving the young girl to die, the robot had made the correct logical choice but it wasn’t the right one. The captain of a starship could be faced with similar moral dilemmas on a day to day basis that should never be left to cold heartless logic of a machine, this thinking however may not apply to Smart AIs yet are mostly still restricted by their programming and with always follow orders therefore not possess any true sense of free will.

[Emberblaque]

Just occurred to me how differently the Keyes loop would have turned out had the Iroquois's AI been installed Keyes may have relied on its input and possibly would never have considered the insane manoeuvre he came up with as more than likely it would have advised against it and suggested a more logical and less risky approach if any.

I would not leave an AI in the same command position of a human captain, to give a better explanation to this reasoning if anyone has seen the Will Smith- I Robot movie a flashback scene reveals his character's backstory where he was involved in a car crash and his car is sinking while pinning a second car containing a trapped young girl, both would have perished were it not for a passer-by robot that witnessed the crash and attempted to save them, there was only enough time to save one of them based on their probability of survival and the robot chose Will Smith's character leaving the young girl to die, the robot had made the correct logical choice but it wasn’t the right one. The captain of a starship could be faced with similar moral dilemmas on a day to day basis that should never be left to cold heartless logic of a machine, this thinking however may not apply to Smart AIs yet are mostly still restricted by their programming and with always follow orders therefore not possess any true sense of free will.

I can only hope that Captain Keyes would have followed UNSC protocol and brazenly defied the superintelligent disembodied human mind's analysis and followed his intuition, as he did serving aboard the Pillar of Autumn above Installation 04, and as Captain Cutter did aboard the Spirit of Fire in virtually every dire situation. Remember that Cutter was a by-the-book kind of guy.

[sloosecannon]

I can only hope that Captain Keyes would have followed UNSC protocol and brazenly defied the superintelligent disembodied human mind's analysis and followed his intuition, as he did serving aboard the Pillar of Autumn above Installation 04, and as Captain Cutter did aboard the Spirit of Fire in virtually every dire situation. Remember that Cutter was a by-the-book kind of guy.

superintelligent disembodied human mind... that's a new term

[Zero]

More like a copy of a dead human brain with most of the memories stripped away but retaining some experiences.

Dead human brain? Cortana was cloned with Dr. Halsey's brain and she was far from dead.

[Bornstellaris]

Dead human brain? Cortana was cloned with Dr. Halsey's brain and she was far from dead.

That was the very first instance where they used a live brain instead of a dead one.

[KhevaKins]

Dead human brain? Cortana was cloned with Dr. Halsey's brain and she was far from dead.

 

 

That was the very first instance where they used a live brain instead of a dead one.

Halsey's flash clone's brain. And I am pretty sure the clone died. That is why it was so controversial.

 

"Cognitive Impression Modeling is a technique used by human scientists to create Smart artificial intelligences by scanning a human brain, referred to as a donor brain. As the brain is destroyed in the process, CIM will only be used on the recently deceased or on a flash clone of a living human^."

[Moustachio86]

I would not leave an AI in the same command position of a human captain, to give a better explanation to this reasoning if anyone has seen the Will Smith- I Robot movie a flashback scene reveals his character's backstory where he was involved in a car crash and his car is sinking while pinning a second car containing a trapped young girl, both would have perished were it not for a passer-by robot that witnessed the crash and attempted to save them, there was only enough time to save one of them based on their probability of survival and the robot chose Will Smith's character leaving the young girl to die, the robot had made the correct logical choice but it wasn’t the right one. The captain of a starship could be faced with similar moral dilemmas on a day to day basis that should never be left to cold heartless logic of a machine, this thinking however may not apply to Smart AIs yet are mostly still restricted by their programming and with always follow orders therefore not possess any true sense of free will.

I was following exactly the same logic, to the point that I was considering using that exact example. The only flaw as I see it is that the robot in question likely could not be compared to a smart A.I. in terms of logic or even compassion. A.I. such as Cortana (maybe a bad example, I think she was supposed to be special in more ways than we know so far) I suspect would be more likely to make the 'right' choice on occasions rather than the logical one. Doesn't mean at all that I'd necessarily trust them in every situation, though.

[spartan-G257]

wasn't it Halsey who also managed to get clones to have the same memories as the previous person (which resulted in bad illnesses to the clone)... if this was the case then wouldn't they also be able to allow the previous memories of the human being to be transferred into the AI as well...

 

i just realized that if they did do that than thy would be better than the forerunners in AI technology (the forerunners failed at doing so (remember the composer))

[KhevaKins]

The Forerunner monitors are based on living tissue and retain their memories.
I don't think the UNSC are unable to leave the AI with memories, I think they don't want to. Say the donor had a grudge while still alive, now they have all the means they could ever need to cause mischief. It would also be harder to control the AI's if they had their owner agenda's from the get go. The UNSC "Smart" AIs do still remember things from the donors life however.

[spartan-G257]

ah very true

[Moustachio86]

i just realized that if they did do that than thy would be better than the forerunners in AI technology (the forerunners failed at doing so (remember the composer))

I'm not sure if I'd go that far. I can't believe Human A.I.s didn't pale in comparison to something like Mecendant or Offensive Bias. Or even the Monitors. That's not to mention the fact Human A.I.s go rampant after 7 years, Forerunner ones are (almost) fully operational after some 10,000 years.

[spartan-G257]

not so much about that but the forerunners tried to do the same thing with the composer and failed

[Moustachio86]

Composer was never a dedicated AI creation method though, it just so happened creating AI copies was the only way to transfer biological tissues to digital entities. In that sense you're right but I would never view the composer as the pinnacle of forerunner AI's or compare its results to Smart AI. Though I admit they have similar origins, their purposes and qualities were entirely different.

[KhevaKins]

not so much about that but the forerunners tried to do the same thing with the composer and failed

The Composer was a way to rapidly transfer physical consciences into digital versions, unaltered, then be able to transfer them back into physical (biological was intended) form. The problem was they kept breaking down and when they attempt to transfer them back into physical form the process kept failing, creating mutants instead. It isn't the same thing as creating an AI.

Guilt Spark is a direct copy of his physical self (Chakas) and he operates just fine, on par or even better then Human AI (on the  UNSC Rubicon he took full control of the ship with the human AI not even knowing). I don't know if Smart AI not keeping the memories is intended or not but if it isn't (i.e. they are supposed to keep there memories) then the Forerunner AI creation methods far exceed human methods.

Also the Contender AI are Jesus level epic, nothing compares.

[Moustachio86]

All of this^

If you need proof of the sheer ability of contender-class AIs, look at the final battle of the Forerunner-Flood war. Mecendant vs. Offensive Bias, the scale is virtually incomprehensible.

[Defender0]

If you need proof of the sheer ability of contender-class AIs, look at the final battle of the Forerunner-Flood war. Mecendant vs. Offensive Bias, the scale is virtually incomprehensible.

ive read the summary of the battle that offensive bias made on halopedia, forerunner AI are merciless when they are trying to achieve victory.