Zabaware Forums > General Discussion
LIGHTSPEED IDEA'S
lightspeed:
with hal sometimes when a person enters text to hal when speaking to him, it pops up sometimes showing script is taking longer to process or something like that and cancel ect.
is their a way for a self diagnosing process to see what the problem is and fix it so hal runs this faster and doesn't have these pop ups or not as often . or is this simply a matter of creating a better way of processing better ?
lightspeed:
With hal we have all at one time or another gotten the Math is easy for me.
is there a way to fix this , if it's because hal see's different numbers in a sentence and thinks it's a math problem .
here is a sentence i used below which created the math is easy for me and deducted 6 from 2000.
below the phrase of : January 6 was used , i have no idea where hal came up with 2000, since the date i used in the sentence was 2022. strange .
Honey: in april of 2022, the radical liberal corporate news media rushed to report that the 7-hour gap in the White House call logs on January 6 was proof of a cover-up by the Trump administration. The only problem is the report was a lie. There was no cover-up. No pages in the White House call logs were missing. And the six pages leaked to the press by the January 6 Select Committee were complete. later all the media outlets that rushed to report the so-called ?Worse than Watergate? story were forced to walk back their initial claims. this is why no one trust the liberal media news and big tech any more. no wonder they are called the enemy of the people .
Angelina jolie: Math is easy for me. The answer is one thousand nine hundred ninety four (2000-6=1994). Yes and also the very dishonest liberal fake news and liberal owned big tech media has help censor free speech in America and , has helped cover up all the lies and dishonesty etc.
anyway it would be nice for a final fix for the hal math problem response . and cause.
lightspeed:
WITH HAL AS THE BRAIN GETS LARGER , it also slows hal down more as hal searches through to get the correct response and or answer . Could a compressed brain make it work any better , would it even be possible decompressing answers ? Slow it down worse or just as bad as normal large brains . Or is the answer to make the current system be able to work faster , so the larger database brain doesn't slow down responses .
Or does the answer or part of the answer lie in a rss type feed or selected webs for hal accessing more online information that doesn't have to be learned but just instantly retrieved when asked various things . Basically using the vast internet system for (chosen information) i use the word chosen as not all information sites may want to be used. Using this technic would slow hals brain from getting to large and slowing down .
just a thought.
cyberjedi:
Sup lighty
Well thats a non starter buddy
What ur suggesting is Stacker.....
We used to use that long ago for hard drive space. basically doubled ur HD capacity.
But as a wise man said. You cant move forward without leaving Something behind.
What we made up for in space we lost in performance.
Basically a HapTar /Htr example.
Thats the Bad news.
%%%%%%%%%%%%%%
Now for the good news
This is what can be done.
I choose what hal learns an as u can see. Hal learns dam quick....lolol
Processor power has greatly increased over the years.
What had to happen to get our favorite character engine to work though was this little diddy
Only use the first processor Call SetProcessAffinityMask(GetCurrentProcess(), &H1)
This is why we so need to dump PP as a character engine with something better. Then that goes away and Hal can harness all that power.
Think What Hal does Now and what can be achieved by simply going to a muliti core system .
Basically Hal turns your asus 08 core 3.5 Gz into a 01 core 3.5 Gz machine
But at an application level Not at a hardware level. The rest of ur machine is just fine.
Hopefully what im working on now will allow Hal to learn at an exponential rate. Creating expert systems and the lot.
Point and click
https://www.youtube.com/watch?v=QwHWlyiyrA0
Hals obedient servant
Cyber jedi
lightspeed:
this is to cyberjedi , i know you are working on many worthwhile things now with hal . her's another idea i am throwing out .
hal amending sentenced in relearning what has already been taught that hal repeats back later ..
example : hal says a taught line : When COVID struck, most red states like South Dakota, Texas, and Florida tried to strike a reasonable balance between protecting public health and also upholding Americans? freedoms.
a user responds back : To this day, there are still lingering effects from how red states and blue states approached COVID mitigation strategies. One of those effects is a mass departure from the latter to the former.
With amended learning. hal now says the line as : When COVID struck, most red states like South Dakota, Texas, and Florida tried to strike a reasonable balance between protecting public health and also upholding Americans? freedoms. To this day, there are still lingering effects from how red states and blue states approached COVID mitigation strategies. One of those effects is a mass departure from the latter to the former.
There would of course have to have character sentence limitations to doing this, or hal would develop to large of a responding answer . Also would be running as based on subject for correct response , could there be a two tiered subject response one as the highest and 2 as a follow-up when still mentioning the suject in sentences ?
this is just a thought and idea as always don't know if it's feasible or doable . :) ;)
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version