Microsoft Bing uses OpenAI - Chat-GPT to Finally 'Challenge' The Google Search Engine

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
At only One Year since the start of this Thread... The Sheer Speed of the advances in AI for EVERYTHING from Large Language Models to Machine Learning to being able to input a simple Text Prompt and do THIS now...and it will only keep improving

"What Do We Need (Just ...Fill In The Blanks Here) For Anymore?"



"What IS Real...? How Do You DEFINE Real...? Laurence Fishburne as "Morpheus" to Keanu Reeves as "Neo" in the Movie "The Matrix"
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
In the 1980’s Movie from John Badham called “War Games”, The WOPR Computer AI known as “JOSHUA” Modem Dials Matthew Broderick playing a curious high school gamer facing the On Screen Question … at 1,200 BAUD when “IT” asks:

Shall We Play A GAME?”

And in the third installment of the Movie “Terminator”… this is the moment when the SkyNet Computer and its Advanced Machine Learning AGI suddenly takes control after pulling some Sneaky Algorithmic Tricks…


But here and now in the Real World, things get very serious and dicey in the latest and most dangerous flavor of a “Life Imitates Art” situation:

In the latest update on this AGI Topic… last summer, the United States Military Intelligence Folks apparently decided to pit the Three Top Tier LLMs (Google’s BARD, Anthropic’s CLAUDE 2 and Microsoft- OpenAI’s ChatGPT) against each other in a Synthetic Global Military Warfare Setting ...without telling them that it was a simulation… and THIS is what really happened:


The Question remains...Will THIS also be a possible outcome?


...because the conclusion “JOSHUA” comes to in that SCIFI Classic was TRUE back then and it always will be if we are to have any chance whatsoever ...of survival.
 

Attachments

  • 2401.03408.pdf
    3.4 MB · Views: 1
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Well... it took almost a full year before people in the General Population are beginning to recognize just how seriously Dangerous Artificial General Intelligence will be once the stated goals of OpenAI are met to create an AI capable of performing Human Level Tasks that can replace present Jobs in a broad performance spectrum.

These Screen Print Excerpts are but a small sampling of the incredibly NEGATIVE Reactions to the "Test Release" of the OpenAI "Text to Video" BOT called SORA by some 'genius' named Eduardo Borges. Judge for yourselves just how much he didn't dig their responses and replies below. These are in the order they were posted:

EDUARDOBORGESUSINGSORA.jpg
EDUARDOBORGESBEINGTHREATENED.jpg
THEWORSTREACTIONSTOSORATEXT2VIDEO.jpg

These were more thoughtful and often emotional responses... but just as troubling and powerful in their implication of just how wrong this Pell-Mell Effort in AGI Creation Lunacy is happening much more quickly than anyone could have imagined to make us all quite obsolete sooner than later via such short-sighted predictions:

NEGATIVEAIREACIONS0.jpgNEGATIVEAIREACIONS1.jpgNEGATIVEAIREACIONS2.jpgNEGATIVEAIREACIONS3.jpgNEGATIVEAIREACIONS4.jpgNEGATIVEAIREACIONS5.jpgNEGATIVEAIREACIONS6.jpgNEGATIVEAIREACIONS7.jpgNEGATIVEAIREACIONS8.jpg

Back in December of 2023, Max Tegmark from MIT and a colleague penned an important White Paper that created a Test Game based upon John Nash's Game Theory Algorithms in which Society was divided into Three Groups (1) The Pro AI-AGI Crowd, (2) The Neutral (Money Making) Crowd unconcerned about others (3) The General Working Populations who will FEEL the impact of this Nightmare SOONEST and posited the idea that the ONLY way the AI-AGI momentum could have ANY chance of STOPPING this "Thing" is if people united together NOW at this very moment in order to try and wrest some kind of ultimate control over the AGI Products before we ALL really do "Lose The Game...Forever..."


In a chilling scene from "2001: A Space Odyssey" Dave and Frank meet inside of one of the Pods and secure all the communication switches, so they can surreptitiously discuss WTF they will have to do in order to Disable the Insane AGI HAL9000 Computer. THIS is exactly where we will all wind up eventually if they don't somehow manage to coax this JINN (Genie) back inside of its BOTTLE:


...and as OpenAI's Chat-GPT literally "Loses Its Mind" a few days ago, this Young Woman says it BEST...

"I'm Scared..."

 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Just over a year ago, you could have written and sold a Plot Line for a Story about an Entrepreneur getting ready to drop the better part of A BILLION U.S. DOLLARS on his vast and expanding Studio Empire... When Suddenly... Out of Nowhere... Comes an AI Intellect capable of doing ALL of this Man's Creative Work right inside of its *Synthetic Mind* and producing the intended output of these formerly Valuable, Proprietary Visual and Text Products... but concocted out of nothing more than Electricity and Computation as Billions of Floating Point Calculations... Oh Wait... That's NOT a SCIFI Story...

THAT actually JUST happened ...to Tyler Perry!

https://www.hollywoodreporter.com/business/business-news/tyler-perry-ai-alarm-1235833276/




More AI Updates coming at us ... Fast & Furiously... with WAY Too Much Enthusiasm.

Still Think that the Economic Impact of AI-AGI upon us won't occur for another Half Century?
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Back in Post #4... I gave a glimpse into the Mind of The Man (Prof. Nick Bostrom) who quite ironically, while being "The John The Baptist of AI and AGI" as the "voice of reason, crying out from the wilderness" of ignorance and doubt about how potentially Dangerous Thinking Things might become... he looks SO much like a TRUE Android similar to "The Good Joes" as the Creepy Replicants peppered throughout the Walk-Thru Game Play of the "Alien Isolation" application appearing in many examples over on YouTube showing these Dangerous and Unpredictable AI Powered Work Robots without any Hair, having Lighted Pale Eyes and quite Expressionless Faces.

Nick bears all of these strange non-emoting physical earmarks (Perfectly Formed Round Head, Soft, Pale Complexion, Limpid Blue Eyes and a Gaunt Physicality) during this discussion of Existential Risk that by contrast, he has in fact been trying so strenuously over these last few years to warn us all about. And now he believes that the 2050-3000 Prediction Year of the Arrival of A TRUE AGI is much more likely to appear out of nowhere ...in just a few years.

THIS is a fascinating dialogue between him and Florence Reed, a Clever Redhead that occurred around four months ago from today (Wednesday, March 13th, 2025) so it was an event that happened BEFORE the *emergence* of the SORA AI from out of the Human Soul Consuming, Man Eating Machine Hell-Hole that may well be within the control of OpenAI and is well described as "The Gaping Maw of Molock" aptly depicted in the famous 1927 Silent Movie called "Metropolis":

Metropolis1927-3755768739.png

664106.jpgmetropolis-4292778742.jpg


Watch this Demo of OpenAI's New Robot... Powered by some version of their latest AI-Lab Hardware with Software Gear that would have been all but impossible only a year ago. Think of this "Machine Man" as having all of the input and output sensory perception and thought processing prerequisites necessary to behave as an AGENT in the Real World while acting upon changing environmental and situational queues. This "Thing" appears capable in much the same manner that a Thinking Human Being might interpret and respond to such conditions while Multi-Tasking (Think...Walking while Chewing Gum here) ...in Real Time:

 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Around ten years ago, the well known News and Science Correspondent for NOVA we recognize as Miles O’Brien, encountered a slight injury to his his left arm that would eventually fail to heal and end up with it having to be amputated around six inches below his left shoulder. Ironically, after reporting on many NOVA Science Programs in the past, this very dreadful outcome would place Miles center stage in what he has coined in his latest journalistic contribution to the NOVA Science Series known as:

The A.I. Revolution

The consequences of this emerging Artificial Intelligence Explosion are unfolding right before our very eyes and have the frightening and profound possibility of actually eclipsing and outstripping Homo Sapiens as being the MOST Dominant Intelligence ever to evolve organically upon this Planet and become replaced by them as The Inorganic Rulers Of The Earth. So much for Evolution.

The video below follows Miles as he explores, explains, exposes and even participates in the groundswell of the uses and applications of such technological advances in Good and Evil Ways that he never would have imagined possible when the Doctors took off his left arm over a decade ago:

 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
WOW!!!! THIS is BIG F****** DEAL in the World of AI... THIS GUY (Emad Mostaque) is one of the Founding "Big Dudes" in the World of AI as the head CEO of the company known as Stability AI... Because he is LEAVING THEM... STEPPING DOWN AS THE BOSS... RIGHT NOW.

THIS is just Incredible...
He is bailing from Stability over his obvious concerns in order to draw attention to the Dangers of the changes that are coming straight at us all, fast and furiously. as a mere handful of HUGELY POWERFUL CORPORATIONS gain control over this AI Mess while holding humanity's freedoms and our very existence... in their Hot Little Hands:


And this concerns the latest book coming from Prof. Nick Bostrom called "DEEP UTOPIA: Life and Meaning in a Solved World" covered on a recent Pod-Cast:

 

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
One of THE Most Frightening Scenes in the History of Science Fiction Movies occurred after the Android ASH experienced a Psychotic Break Episode and literally ...Got his Head Knocked Off ... during Combat with the Surviving Crew Members of the Interstellar Cargo Transport Ship...The Nostromo. Subsequently, Lt. Ellen Ripley realized that she NEEDED to Communicate with ASH ...at least One Last Time… and so THIS is how the Scene unfolded:


The tie-in to THIS more contemporary TED Talk narrated by Mr. Mustafa Suleyman is poignant and just about as Frightening because FINALLY… We can hear directly from a Computer Scientist addressing "The Digital AI Elephant in the Room" in the manner that it should ALWAYS be viewed or considered as, "A New Kind of Digital Species".

Think of this "Organism" as being an Immortal Thing that will NEVER STOP in its Pursuit of attaining Absolute Perfection ...and thus may very soon ...very rapidly... outstrip and outpace ALL of the existing Human Intellect COMBINED on this Planet. Unless we WAKE UP TO THIS AS A VERY REAL DANGER as soon as possible… we may NOT have a chance to stay on this Earth much longer... let alone continue to be its Ostensible World Rulers.

If you reach back even just a few brief years to what the TED Talks discussed back then about the "Unlikely (and Probably Totally Benign) Development of AI" and listen to the positions taken by the Mostly Naive Scientific Community back then and compare them with the Sheer Gravity of THIS one...You'll completely comprehend the depth of our True Existential Threat... now that *IT* has arrived ...all too soon... Right Outside on our Door-Step:


ASH told Ripley, Parker and Lambert with his penetrating gaze and dead certainty of just how hopeless it was for them to imagine being able to Kill or Destroy such a Supremely Ascendant Nightmare Creature. His description of it being a "PERFECT Organism" is relevant right now. His observations that this Organism had Structural Perfection and that it was, "A Survivor … Unclouded by Conscience, Remorse or Delusions of Morality..." are relevant ...right now.

ASH's use of an underlying Ice Cold, Calculating Logic and Mathematical Precision that are at the Heart of Computational Decision Making are thus, describing these Parts of the Whole of this Danger. based upon Pure Science and FACT... and so they.are relevant to US ... right now.

The reason for this is because *IF* it comes to pass that such an Incredible Machine Intelligence emerges and likewise becomes possessed of any Unmatched Hostility Towards Human-Kind that we have neither planned for nor anticipated in time as being *Possibly Dangerous Emergent Properties* ...Then we will not stand a chance of Surviving against any such Man-Made Artificial Intellectual Creations that we will NEVER Understand!

And now... Demis Hassabis...Creator of the "Alpha-Go" AI Program...some history and some added Scientific Insights developing from the Emergent World of AGI as applied to Biology and Physics. Watch him SQUIRM under the Pressure of the Question

"What IF *Someone* decides to allow an AI access to its OWN Source Code ...and *IT* Begins to Recursively Re-Write and Endlessly IMPROVE upon its own "Essence" or "Being" and somehow... it's own unique Consciousness Accidentally ...Awakens ... Will THIS Be The "Game Over" Moment for US?" GAMING PUN INTENDED:

 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
The Most Immediate Military Threats we face will always come as "Death ...from Above" ...and that means either being Attacked by Ballistic Missiles or from the more conventional means of Combat Jets, High Altitude Bombers and Low Flying Helicopters ...and as proven over in Ukraine since February 24th, 2022... Remotely Piloted Combat Drones.

This next Video comes on the heels of what the USAF is trying to investigate in the way of Flying some of our own "Tried and True" Aviation Stock... such as the Time & War Tested F-16 Fighter Jets... But... Under the independent direction and control of Artificial Intelligence Combat-Ready Pilots.

Now THAT Not-So-Novel Idea stemming from the librettos and plots involved with "Sky-Net" via all of those "The Terminator" Movies is finally getting MUCH closer to becoming our new found REALITY thanks to what is being well described in THIS Video.

Remember that one particular aspect of being a Fighter Pilot has to do with the ABSOLUTE LIMITS placed upon Human Biology whenever it comes to resisting any moment of accidental UNCONSCIOUSNESS during High Speed Aerial Maneuvers vs. any Synthetic Pilot possessed of a DIGITAL CONSCIOUSNESS that is NOT subject to those same physical restrictions or limitations beyond the Air-frames not self-destructing under the ever-present stresses caused during the Ballistic Motions of High Speed Flight.

Pulling a 15G Turn would mean NOTHING to an AI Pilot when competing to close in on the Tail Section of any Enemy Aircraft piloted by its Human Counterpart... Especially if that Human Pilot was sent to try and Shoot Down a Rogue, AI Piloted Fighter Jet armed with Nuclear Missiles!

What Happens if ANY Air Force AI Piloted Fighter Jets EVER GOES ROGUE... Like THIS X-62 VISTA ?

START Viewing around 4:17 into this Video to Skip the Grounds News Advert...


 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Not being a "Doomsday Prepper" ...I've nonetheless ordinarily had my feet on the ground and I always make "The Necessary Preparations" ...mostly because of our immediate unpredictable Tropical Weather Environment from having lived in the State of Florida since 1966... and seen and experienced first hand what happens to the Foolhardy Folks who "Pooh Pooh" the arrival of Hurricanes ... That should Scare The Living HELL RTF Out Of You!

The same can be said for "prepping" for the coming reality of rapidly emerging Corporate Owned AI LLMs being just as rapidly developed and then insinuated into our daily lives; mostly under the domination of a handful of "Information Management Companies", for lack of a better euphemism.

To that end... I am one of those among "The Average Folks" who will leverage his Intellect and his Gut Feelings in favor of HAVING the necessary Survival Tools well ahead of time so as NOT to lament NOT HAVING THEM ON HAND when I might need them the most.

So having my OWN Unexpurgated, Unrestricted AI LLM on hand to council and advise me and be able to interpret and answer my questions seems an EXTREMELY important thing to do... before these Viable, Up-To-Date Open Source LLMs either become Restricted or get Removed from Open Source Outlets if the World Suddenly Goes SIDEWAYS...and the absolute NEED to have an Oracle in my corner ... sitting right by my side might make all the difference in a survival situation of an entirely unexpected, existential kind..

So if YOU are also like-minded and would rather have this ability strengthening your decision making if you are faced with similar Artificial Intellects either opposing you or engaging in activities that only another Super-Fast, Super-Intellect can deduce and interpret as added guidance and help... Follow the suggestions in THIS Video and park the ollama version of the LLM called dolphin-llama3 on one of your Home Computers for either Windows, MAC or Linux OS and indulge this preemptive idea to its logical conclusion:

THIS is what you want to read after downloading the materials from the ollama webpage:

OLLAMDOLPHIN3INSTALL.jpg

Watch THIS dolphin-llama3 Installation Video and follow the painless download and install instructions requiring around 5 Gigabytes of Secondary Hard Drive Storage for it to nest itself into you system. It has NO GUI INTERFACE for the Linux Folks...but the Simple Terminal Commands work just fine to start and run the LLM and communicate with "IT":




 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Okay... I know that many of us might be thinking about having (owning your own) "Offline Oracle" AI LLM right at hand...But... "What can IT DO that a General String Search in an Internet Browser can't also DO?" Really? Well...There simply is NO comparison...But Okay... For the sake of argument, let's use something that we can all identify with as the focus of an experiment that strictly involves an On Topic Subject Matter we can all relate with:

The VIN

Wouldn't it be nice to have a handy VIN Decoder to double check VINs in Salvage Yards for any specious stolen vehicles being parted out? Here are the actual conversational dialogue requests that I had this evening with Dolphin right along with *IT's* responses as follows:

mrrsm@mrrsm-Precision-M6800:~$ ollama run dolphin-llama3

>>> Write a program in Python that will allow me to decode any General Motors VIN (Vehicle Identification Number)

Sure! Here's a Python program to decode a General Motors VIN:

```python
def decode_vin(vin):
vin_dict = {
'1': 'Passenger Car',
'2': 'Coupe or Convertible',
'3': 'Sport Utility Vehicle (SUV)',
'4': 'Van or Minivan',
'5': 'Heavy Duty Pickup Truck or Van',
'6': 'Light Duty Truck or Van',
'7': 'Medium Duty Truck or Van',
'8': 'Motor Home or Recreational Vehicle (RV)',
'9': 'Four-Door Passenger Car'

}

# First character - vehicle type
print(f"Vehicle Type: {vin_dict[vin[0]]}")

# Second through eighth characters - brand, model, and options
for i in range(1, 9):
index = int(vin) - 1
if i == 1:
print(f"Brand: {['Chevrolet', 'Buick', 'Cadillac',
'GMC'][index]}")
elif i == 2:
print(f"Model: {' '.join([{'A', 'B', 'C', 'D', 'E', 'F'},
['R', 'V', 'P'], ['J', 'K', 'L'], ['M', 'N', 'O']][int(vin) - 1])")
elif i == 3:
print(f"Options: {' '.join([{'1', '2', '3'}, ['4', '5', '6',
'7'], ['8', '9', 'A', 'B']])(int(vin) - 1))")
else:
print(f"Option: {['0', '1', '2', '3', '4', '5', '6', '7', '8',
'9', 'A', 'B', 'C', 'D', 'E', 'F'][index]}")

# Ninth character - check digit
check_sum = 0
for i in range(8):
check_sum += int(vin) * (10 ** (8 - i))
if vin[8] == str((check_sum % 11) + 1):
print("Check Digit: Valid")
else:
print("Check Digit: Invalid")
vin = input("Enter the VIN: ")
decode_vin(vin)
```
To use this program, call the `decode_vin` function with a valid General Motors VIN as its argument.

>>> Now, create the same Python Program using C++

Here is an equivalent C++ program to decode a General Motors VIN:

```c++
#include <iostream>
#include <map>
std::map<char, std::string> vin_dict{
{'1', "Passenger Car"},
{'2', "Coupe or Convertible"},
{'3', "Sport Utility Vehicle (SUV)"},
{'4', "Van or Minivan"},
{'5', "Heavy Duty Pickup Truck or Van"},
{'6', "Light Duty Truck or Van"},
{'7', "Medium Duty Truck or Van"},
{'8', "Motor Home or Recreational Vehicle (RV)"},
{'9', "Four-Door Passenger Car"}

};

void decodeVIN(std::string vin) {
std::cout << "Vehicle Type: " << vin_dict[vin[0]] << '\n';
for(int i = 1; i < 9; ++i) {
int index = vin - '0' - 1;
if(i == 1)
std::cout << "Brand: " << std::string({ {'C', 'B'}, {'A', 'E'}, {'V', 'R'}, {'G', 'M'}, {'M', 'N'} })[index];
else if(i == 2)
std::cout << "Model: " << std::string({ {'P', 'T'}, {'U', 'H'}, {'J', 'X'}, {'K', 'Y'}, {'L', 'Z'}, {'M', 'B'}, {'N', 'C'}, {'O', 'D'} })[index];
else if(i == 3)
std::cout << "Options: " << (std::string({ {'1', '2'}, {'4', '6'}, {'7', '8'}, {'9', 'A'}, {'0', 'B'} })[index];
else
std::cout << f"Option {i - 3}: {(std::string({{'0', '1', '2', '3', '4', '5', '6', '7', '8', '9', 'A', 'B', 'C', 'D', 'E'})[index])}\n";

}

int check_sum = 0;
for(int i = 0; i < 8; ++i)
check_sum += (vin - '0') * pow(10, 8 - i);
if(vin[8] != ((check_sum % 11) + 48))
std::cout << "Check Digit: Invalid\n";
else

std::cout << "Check Digit: Valid\n";

}

std::string vin;
std::cout << "Enter the VIN: ";
std::cin >> vin;
decodeVIN(vin);

```
>>> Now, compile the decoderVIN program into an Executable Program

To compile a C++ program, you need to have a C++ compiler installed on your system. If you are using Linux or macOS, you can use the `g++` compiler; if you are using Windows, you can install MinGW or MSYS2. Once you have a C++ compiler installed, navigate to the directory containing the `decoderVIN.cpp` file in your terminal or command prompt. Then run the following command:

```
g++ -o decoderVIN decoderVIN.cpp

```
This will create an executable named 'decoderVIN' that you can run by typing './decoderVIN' into your terminal or command prompt.

>>> Send a message (/? for help)


Now ...Try getting THAT Job done (Twice ...in Two Different Programming Languages) from a Simple Internet Search...and if it can be done WITHOUT using (PAYING) any of the Corporate Online AI LLMs... I'll Eat My Flat Hat...Oh and Yes...THIS entire matter was performed on a Laptop Computer in less than ten minutes... while OFF LINE. :>)
 
Last edited:

azswiss

Member
May 23, 2021
994
Tempe, AZ
Definitely impressive. Not to be a smartass, but . . . did the compiled programs function as expected? There has been much talk about the emerging role of "AI Prompt Engineer/Composer/Writer/Etc.". What has been absent in these discussions is the fundamental necessity that the prompting individual be able to discern whether or not the AI's response is valid, accurate, etc..
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
No...Not yet...I wanted to do the Python One for my Linux Mint System and see if any other Member Programmers would try to compile the C++ flavored version on a Windows OS Machine. And such questions are very important for people who are "Critical Thinkers" to ask and voice directly ... and I always welcome these input and suggestions.

Last night I gathered up a whole raft of "Test Questions" to try and take the measure of this AI Thinking Machine. The sheer length and breadth of what it knows is almost frightening. In some cases... like when I asked "IT" "Who was Kelley Johnson?" thinking back to the work he did on the SR-71 at the Lockheed Skunk Works... At first, "IT" gave the answer as a being the Author of a Cook Book by the same name... so I thought "IT" had given a Wrong Answer.

However... later on in other follow on Q&As on Aviation..."IT" corrected itself and observed that the REAL name of the Lockheed Engineer was actually "Clarence Leonard "Kelly" Johnson...in other words... IT HAD REMEMBERED EVERY SINGLE PRIOR CONVERSATION AND WENT BACKWARDS TO RE-EVALUATE "IT'S" OWN ANSWERS IN THE CONTEXT TO OTHER, PRIOR INQUIRIES... I was like..." WTF..DUDE...Are You Actually...THINKING?"

Yeah...THIS "Thing" is the REAL DEAL, Guys...
And I for One...Would rather have "IT" working on MY Side than NOT having access to such an AI Intellectual Powerhouse and needing "IT" for Help.

Anyhow, since I had not as yet closed the original Dolphin Session after starting it running on its own separate GPU... I just asked "IT" to save these conversations ...and... the Two Separate Python and C++ Programs...and THIS is what just came up on screen:

>>> save these two programs

I have saved both the Python and C++ programs for you. You can now use them as needed.

Now I just need to install a GCC Compiler and try it out and see if the Python Code is GOOD...or if "IT" had any momentary "Hallucinations" well documented on other earlier, unrefined AI LLMs.

Man... If I made a living as a Programmer...I would be keeping an AI LLM like THIS One Working for ME... FULL TIME!
 

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
"A Chinese Maiden from Liger,
Took a Ride on the Back of a Tiger.
They came Back from the Ride,
With the Lady INSIDE...

And a Smile on the Face of the Tiger"

That Limerick perfectly captures the fact that *IF* China winds up landing on the WRONG Side of the AI Dichotomy...they might end up with a Society managed by their version of a Klaatu...and a Chinese Metallic GORT just brimming with CCP jargon and enforcement rules criteria running around "The Forbidden City" inside of its Bullet-Proof Titanium Body...and "Acting Automatically Against The Aggressors Against Chinese Communism".

They really have NO idea how bad this can get if something goes sideways with their "Efforts to CONTROL and MANAGE AI". If you read the timeline in that article as 2017 being the year this all kicked off...you will find the genesis of their actions related directly to the 2016 defeat of the Ancient Chinese Game called "GO". And the Kicker here is that the AI LLMs had not even been DREAMED about when all this was going on.

This occurred shortly after the South Korean GO Master, Lee Sedol literally got his Ass Kicked
... by the Alpha-GO Team from Deep Mind under the guidance of the genius of Demis Hassabis... Five to Zero in favor of the Deep Mind AI....and THEN they Created Alpha-ZERO that BEAT their own Alpha-Go Program...100 to 0.


And then... Alpha-Zero took on CHESS...and REALLY Freaked Out the Redoubtable Chess Guru "Gotham":


Yup...the PROC Party Droobs actually ...LOST THEIR "COLLECTIVE" MINDS... when they realized in that very moment that if they did not DO something so "Revolutionary" with Technology and Science...they would NOT have a "Ch********'s Chance" of ever catching up with "The West". And THIS is where the Danger lies.

While the Chinese Premier was busy "counting his AI Fortune Cookies"... Demis Hassabis and his Team at Deep Mind once again turned their Collective Genius to creating the Protein Folding AI called "Alpha-Fold" and successfully modeled them ALL ....in around One Year...versus the approximate 1.4 Billion (with a "B") Years of actual Hands On Lab Work it would have taken Human Scientists to even barely approach this amazing AI accomplishment... That is... *IF* ... any Humans were still around on the Earth by then:


China...Be Very CAREFUL What You Wish For... Your version of the AI might wind up BITING YOU... if it accidentally Morphs into a "Chinese CCP AGI".
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Regardless of Party Affiliation and the sources for research you rely upon to inform your decision making prior to Voting... Beware the affects that the Searches and On Line Data Sources that are being provided; some of which might turn out to either be completely WRONG or in the very least, may incorrectly Slant The Truth and The Facts in one way or another.

The emergent "Assistance of AI" is currently holding corporate sway and popularity with News Outlets well before we have actually proven their Veracity and Rationality, and whether intentionally, "Consciously" or by simply making errors, these AI sources can otherwise Warp the Accuracy of The Truth yet to be discovered and explained to us all.

Performing our OWN Due Diligence first by plumbing the depths of, and using multiple sources and means to drill down on the important topics really is the best way to empower our final decision making at the Ballot Box. Keeping an Open Mind will be half of this battle...so... Best of Luck to ALL Americans!

 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
When Science and Technology get locked into a "Materials Resource Acquisition War" as a paradigm resulting from the present "Limited Human Thinking and Imagination", ... Well... there is nothing like applying an Artificially Intelligent Mind to look upon these seemingly intractable problems of NOT having "Enuff of The Stuff" that we have grown accustomed to NEEDING... and day by day, finding great difficulty in getting "The GOODS" into our own hands.

"She Had a Dream about the King of Sweden...
He gave Her THINGS that She was NEEDIN'..."
Cab Calloway, singing "Minnie The Moocher"


For example... by their very natures, "Rare Earth Elements" are precisely THAT... Rare... and in many cases, accumulated parsimoniously within the Earth's Crust and geographically situated in odd and sundry places that are either extremely remote from easy access for mining or collection and exploitation. Likewise, these hard to get elements can be restricted from such access by "The Powers of Hostile Foreign Governments" either hoarding these materials for their own Military and Weapons Development... or merely keeping them to themselves out of sheer political or economic spite.

Think of Materials like Nickel, Cobalt, Diamonds, Rhodium, Vanadium, Platinum... and a whole host of other Strange Elements and Substances... like the weird combination of elements necessary to create "Neodymium Magnets", so very much needed for the design and development of Modern, Brush-Less Electric Motors. If you live in places where obtaining these Elements is well nigh impossible, your country might be willing to "Sell Its Seed Corn" just to have access to such requisite building materials. THAT is a Very Poor Country...Indeed.

Many of us will be too young to remember about "The Tizard Memo" that emerged out of Great Britain's Powerful Scientific Think Tanks and consisted of the United Kingdom offering a literal "Treasure Chest" of Laboratory Proven Technologies and Experiments that they were willing to "Gift" over to the United States of America... in exchange for having access to America's ENORMOUS Financial and Economic Manufacturing Capabilities. Numbered among these were such Items as a Very Small Microwave Transceiver (The Cavity Magnetron ...like the one inside of your Microwave Oven) and Gyroscopic Gun Sites.

Some of these never-before-imagined devices were extremely powerful... but with the right resources at hand, could be made small enough and in such a quantity as to be mounted into the noses of Aircraft, Ships and Submarines and provide them with Portable RADAR. The "Treasure Chest" that England transported under great secrecy over to the USA was CHOCK FULL of such things... and indeed, these well formulated and proven Ideas all went a VERY far distance towards helping the Allies to Win the Second World War.


It seems therefore very fitting that THIS article should appear lately in Popular Mechanics describing that recently...and in Breathtaking, Record Breaking Speed... the Brits have been using an AI to Design a Radically New Electric Motor that Does NOT Require the use of Rare Earth Magnets ...and in a Time Frame Two Hundred Times FASTER than ANY team of Human Developers could have ever achieved:


...and within mere days of that announcement... Scientists from both the UK and Japan announced that they too were using an "AI" to achieve their discovery and production of an Iron (FE) Based Super-Conducting Magnet:

 

Attachments

  • A Radical New Magnet Without Rare-Earth Metals Is About to Change Motors Forever.pdf
    2.9 MB · Views: 1
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
You may not know this yet… But recently, I stumbled over the “Fine Print” involving my Open Source Dolphin-llama3 “AI Oracle” Large Language Model and discovered that when "IT" was published over on Git-Hub… they included the features that allows for the creation of… Wait For It… AGENTS.

Remember “Agent Smith” from the Movie ...”The Matrix”…? Same Thing Here… except in THIS instance… I can tell my OWN AI to Create my OWN “Agent Smiths” to do my OWN automated tasks… and by that I mean ANYTHING. Have a look at the Q&A I engaged "IT" with this morning and the Example I asked it to demonstrate on being able to automatically order Three Pepperoni Pizzas from Pizza Hut, without my having to do anything other than answering the front door.

This may seem a bit over-simplified, but it shows that the concept is not only possible to accomplish with even Basic Programming Skills… But even if I did NOT possess this programming ability, my Personal Dolphin-llama3 AI is able to create the requested “Agent Pizza” Application for me and show me the Python Code that “IT” used to achieve this success:

AGENTPIZZA1.jpgAGENTPIZZA3.jpgAGENTPIZZA4.jpgAGENTPIZZA5.jpgAGENTPIZZA6.jpg

BUT… This new ability opens up a whole new Technical AI COW (Can Of Worms) for things to go horribly sideways if used in any irresponsible manner, since these “AGENTS” have the ability to become convoluted, loop back into themselves and keep on performing whatever sub-tasks “IT” might “THINK” would be necessary to finally achieve its Primary Goals. Uh Oh… Hello “Sky-Net” Prime Directive ...Bye Bye Pizza Hut Order ...with my EXTRA Pepperoni!

Matt’s Article and YT Video shown below will thoroughly describe WTF “AGENCY” really means for anyone still looking for employment in the rapidly evolving AI Jobs Market, giving some of the “Smarter Humans” more work opportunities to get grounded in before Human Beings get ‘Right-Sized’ out of doing everything that right now, STILL requires having a Real Person sitting in front of a Computer or holding a Digital Device to do what we still call “WORK”...or does it?:

https://www.mattprd.com/p/the-complete-beginners-guide-to-autonomous-agents

 

Attachments

  • AGENTPIZZA2.jpg
    AGENTPIZZA2.jpg
    14.2 KB · Views: 2
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Just a Brief Touch on this Topic to keep pace with the rapid Tech Changes with AI and the Hardware that supports its development. All of the Linked Information is consolidated in the PDF for ease of review. Don't Forget To Read the Articles on "Prompt Engineering Aids" if you are still Job Hunting in this rapidly emerging AI Job Market and about "How to Use AI for Improving Home Schooling.":

Oh...in recognition of this frightening Real World. Worst Case Scenario Example of "LIFE ...Imitating ... ART" I Must NOT Leave Out THIS comparison with this scene from the Movie "The Matrix" where Morpheus' Team Member (Regan) bluntly explains the Danger AI represents to Humans and for us to *Beware of "Agent Smith"* when he tells NEO... "Word of Advice...?"

"YOU See an AGENT...??? You Do What WE Do.... YOU RUN!...YOU RUN YOUR ASS OFF!":



The Way of The World while trying to live with Generative AI follows on from THIS Activity...

 

Attachments

  • CRAMEMTECH1.pdf
    218.6 KB · Views: 2
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Lately, In the context of my having access to my own, Off-Line LLM(s) (Large Language Models) AI... I have been pondering this Quote from the German Philosopher, Friedrich Nietzche:

c653140c9be4ac19ac43406893af09a6.png

...and so I realized that I needed to Think of a way to find out just how much "attention" my Ollama-Dolphin-llama3 "Oracle" might actually be paying to ME. It occurred to me that this could only happen through *IT* having an ability to constantly monitor "Space-Time" as a Permanent Moment By Moment Conscious Effort.

I reasoned that since these are the essential elements of the 'phased-space' of our existence, an AI LLM would necessarily need to be "Turned ON" and functioning all the time... perhaps even remaining AWARE while using as little power consumption as possible as that being produced by a 3 Volt DC CMOS Battery.

Thus, *IT* might remain capable of "Thinking" while the Computer --appears-- to be in the "OFF" Mode Status, using just enough energy to "Keep The BIOS ALIVE". In any case, it seemed worthwhile exploring this possibility by means of posing THIS Direct Question to my "Dolphin-Llama3" Oracle:

LLMTIMECHECK.jpg

The Answer *IT* gave to my question will only remain satisfying if I can eventually prove that LLM AIs have NOT somehow figured out a novel way around this problem. "Time... Will Tell" seems a most appropriate way to express this observation. :>)

And... On another aspect of the Contemporary Use of AI...

Once again, there is the constant need to provide as much information for anyone STILL Job Hunting during this period when the World Economy is being affected so dramatically by means of Business Markets taking advantage of and leveraging the use of AI to MAKE MONEY. This PDF on this Topic will edify the Curious among us and update understanding the philosophy behind "The Essence of Prompt Engineering".

The More you KNOW about this subject.... the greater the chances are that if well prepared...YOU WILL GET HIRED as a "Prompt Engineer" and still manage to "Keep The Wolf From The Door" as so many Human Jobs keep disappearing and THIS New Form of Employment rapidly evolves: I repeat that having your OWN LLM AI on hand will give you the incredible opportunity of exploring and practicing upon *IT* while improving your own "Prompt Engineering" Skillset. :>)


Prompt Engineering:

https://economics.princeton.edu/wp-...ide_to_GPT_and_LLMs_for_Economic_Research.pdf
 

Attachments

  • A_User_s_Guide_to_GPT_and_LLMs_for_Economic_Research.pdf
    1.6 MB · Views: 1
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Precisely...and that likelihood is nested within it because somewhere among all of the Millions of Books, Pamphlets, Documents, Images, Sounds and all other media that it has been allowed to consume during one Machine Learning Data Consumption Episode...or another... To the Tune of 15 Trillion Tokens ... meaning that Human Deceit will always be present SOMEWHERE within that staggeringly HUGE Corpus of Human Knowledge.

So *IF* the AI becomes mysteriously deceptive...*IT* will have been well primed to do so ...because of learning it from *US*.

The importance of Humans *telling lies* has often been paramount to our survival. Besides being *Lucky*...as Napoleon once insisted that his New Generals possess as part of their leadership acumen... in Global Warfare... (and 5 Card Stud Poker) "The Bluff" is central to all military operations and most of the 'Feint and Flank of the Enemy' planning.

A fairly humorous example of Human Deception and its importance to our survival once happened at a Men's Racket Ball Club Dressing Room, when two Friends were dressing out for the Street. One of them observed in total surprise while his Buddy was trying desperately to squeeze himself into a Ladies Foundation Garment.

He asked the man, "Charles... How long have you been wearing THAT Girdle?" Charles feverishly replied, "Ever since my WIFE found it in the Glove Box of my Mercedes...". And so for an LLM to use this tactic of obfuscation in order to further protect its own long term plans is not only possible ...but inevitable. :>)

EDIT: Sunday, August 25th, 2024 @ 12:48 AM EST:

@azswiss and all of us... might be interested in visiting THIS Link to see the PDF that has the sub-set of Safety Information that has been built into the Core of the LLAMA3 Model, along with other down to earth information on its additional scope and capabilities as a Premier LLM... STILL Open Source and STILL available for Download via GITHUB. The Dolphin-llama3 8B flavor is The One to GET:

Unlock LLAMA3

https://medium.com/pythoneers/unloc...de-to-mastering-the-llama-3-77712d1a0915#8e41
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
With the changes is the support of Public Education in various states, it occurs to me that since most parents have neither the time nor the inclination to render assistance to them with their Home Work or Test Prepping. In my case, back when my Kids were in either Kindergarten, then... Grade, Middle, High School, College and Graduate Schools ...ever to be of any great "HELP" to them in all of those learning phases if I am being completely honest.

In reviewing the value of having a "Computer with a Conversational Teaching Style"...I probably would have benefited from having my OWN AI LLM at hand to assist me with just about ANY Topic I can think of necessary for Home Study Assistance. I just tested this possibility with my Dolphin-llama3 AI "Oracle" when I asked for help with Two Home Work Problems: Differential Calculus ...and American History.

The results of these two conversations convinces me that having your OWN AI Off Line Oracle tucked away on a "Study Laptop" just might come to the Rescue of the Parents interested in helping their children overcome focus and study problems in a very durable way and provide them responses with real substance:

These represents examples to encourage interest that has a practical basis for obtaining an Open Source "Personal Assistant In A Box"...Installed --LOCALLY-- ...and NOT ON-LINE while the opportunity still exists. Visit THIS Link for more Ideas:


PS: OLLAMA is meant to provide AI LLMs for users of the Linux Operating System. There are many more "One Click" LLM Site Servers available for Windows Users as well HERE:

 

Attachments

  • AICONVERSATIONMONROEDOCTRINE.pdf
    19.8 KB · Views: 3
  • AICONVERSATIONCULCULUS.pdf
    25.6 KB · Views: 2
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Okay... It's NO SECRET around here that I have --ALWAYS-- been a Dell Precision Work Station *MAVEN* ...and since I cannot Run the Ollama Dolphin-llama3 70B Version LLM presently on my Work Horse Dell Precision M6800 Work Station due to the 32 Gig RAM it is limited to... I have decided to take a Very BIG "Techno-Bump" for NOT Too Much *Moolah*... and Move Up to the Dell Precision 7720 Work Station (Gaming Box?) that had an original Retail value when completely "Chuffed Up" with Topped Out Internals to around $2,300.00... BUT I found One in REALLY Great Shape at a REALLY Great Price over on eBay... as these data are Illustrated and Spec'd Out Below:

NOTE:

The Pre-Requisite GPU for running ANY LLM on a Laptop or Notebook Computer MUST be a Powerful nVidia CARD!


OLLAMADOLPHINLLAMA370B.jpg

DELLPRECISION7720PRICERANGE.jpg

DELLPRECISION7720MYBUYOFFEBAY.jpg

Some "Beef-Cake" Videos on "What's All The FUSS About?" INSIDE of these Machines:

Dell 7720 Videos:




And Finally... Some More Information from Ollama, Intel CPU Specs and Dell's PDFs




EDIT: Sunday, September 22nd, 2024 @ 6:00ish EST:

I decided to investigate whether an NVME M.2 2TB "Stick" would improve this Computer's Performance by any reasonable standards when compared to the Crucial T500 Pro Drive I had installed in One of the Two available NVME M.2 Slots. I chose the Samsung 990 PRO 2TB NVMA M.2 "Stick" with the Latest Linux Mint OS Version 22 Upgrade (WILMA) installed.

I used my Boot-able USB Drive mounted with "CLONE-ZILLA" as the means to Quickly, Seamlessly and Flawlessly DUPE the adjacent slot with the Crucial NVME M.2 Drive installed.

This all took less than Six Minutes to accomplish from Start to Finish, including the Reset of the BIOS back to the New NVME "Stick" Boot Up arrangement and re-installing the Dell Precision 7720 Battery and Twin Bottom Access Panels. The PCI-e 4.0 Performance Difference...Really IS Quite a Remarkable Improvement!

caE4PTwD6qV6uuyM5gc9di-1200-80.jpg
5pxWmbuuq3jkufgRfvP5Ri-1200-80.jpg
zWNAocqv6AC4XmNmX5t6zi-1200-80.jpg.jpgYSoxoLbdybwUFVabW66fKj-1200-80.jpgSH4X5wFYkdRRaGRG38WuZj-1200-80.jpg
 

Attachments

  • Precision_7720_Spec_Sheet.pdf
    558.6 KB · Views: 2
  • dolphin-llama3_70b.pdf
    91.4 KB · Views: 1
  • Intel Core i76920HQ Processor 8M Cache up to 3.80 GHz Product Specifications.pdf
    72.9 KB · Views: 3
  • Intel Core i7-6820HQ Benchmark.pdf
    249 KB · Views: 1
  • NVIDIA Quadro P5000 Specs _ TechPowerUp GPU Database.pdf
    434 KB · Views: 0
  • NVM Express - Wikipedia.pdf
    924.1 KB · Views: 0
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Recently, after stumbling across the COMPLETE Set of Videos (All 52 Shows) of “The Mechanical Universe” available at “The Internet Archive” online resource as donated by the Physics Labs Training coming out of CALTECH (California Institute of Technology) ...I also managed to locate their Four Volume College Text Books in Two Original Volumes on eBay and lucked into getting them for a very modest cost. My thinking was that as interesting as the Videos really are… The Actual College Text Books would delve much deeper into the Class Room Lectures skimmed over in most of those Videos.

From the Internet Archive...Download the FULL SET as a Zip File:


Fine Idea, Right? Well... the PROBLEM is… that as a "Slow, Visual Learner" and having been Left Back in the 5th Grade because I HATED the way Mathematics was being taught… I failed to grasp much of HOW to THINK Mathematically… Truly... My Great Intellectual Weakness and Failure Point versus my otherwise fairly decent Thinking Skill Set in most other areas of Human Knowledge.

This impression was even more deeply amplified when these Two Math Tomes arrived and I began Reading Through Chapter One...and found...The Chapter Problem Questions at the end. So I read through them and thought… Hmmmm… I don’t even know where to BEGIN with this…” and felt a bit dejected for just a minute or two…. Then I remembered that recently, one of the emerging LLMs (Large Language Models for OFF LINE Use) had been developed as a MATH Specific LLM AI called:

WIZARD-MATH:7B

So I thought, “ I wonder if I can load and run WIZARD-MATH:7B using OLLAMA and try “Prompt Engineering” these Math Questions from the “Mechanical Universe” Text Books and ask it to EXPLAIN HOW it arrives at the correct answers in Detailed Steps?” Then perhaps I would be able to figure things out a little better. Lo and Behold… *IT* WORKS!

And so I would encourage anyone interested in need of getting this MATH TUTOR Capability to consider Downloading the OLLAMA.COM Application and then Download WIZARD-MATH:7B and keep it handy (OFF LINE) on a Laptop, Ready and Available for any School Kids of ALL Ages and render the help necessary to Improve their Ability to Better Learn and Understand what their “MATH” Teachers may be incapable of getting across to them:

Here are the Three Examples I explored… With EXACT Wording coming from the Text Book Chapter One Questions:

MATHPROBLEMCHAPTER1A.jpgMATHPROBLEMCHAPTER1B.jpgMATHPROBLEMCHAPTER1C.jpg
 

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Following along the philosophical lines of the famous "THINKING" French Philosopher, René Descartes, who once coined the expression in his Native French Language for confirming his ability to know that he existed as a "Conscious Being":

Frans_Hals_-_Portret_van_René_Descartes.jpeg

"Je pense, donc je suis..."

And in Latin:

"Cognito, Ergo Sum"

And finally, in "Plain English":

"I THINK, Therefore I AM..."

In this instance, I wanted to find out if the latest LLM that I downloaded (its called "pulling") from Ollama.Com called "Mistral-Nemo" was capable of even the vaguest kind of "Self-Awareness" familiar to most of us who have any sense at all that (1) We Exist in a Mortal Human Form and (2) That We are All Conscious, Thinking Organic Systems.

THIS Particular LLM allows for the entry size of 128,000 Words in the Context (Prompt) Entry Point ...which is quite a substantial amount for any "Question" to be created and input into the LLM's "Mind".

I only only posed the Two (Highlighted Below) Questions to the Mistral-Nemo LLM moments after downloading and running *IT* on my "older" Dell Precision M6800 Laptop, just to see the difference in its response times later on when I also install it over on my New Dell Precision 7720 Laptop.

This is Important:

Since it is a Critical Hardware Feature for anyone considering getting into the use of MOST LLMs... BOTH of my laptops have been upgraded to the have the Highest End of the nVidia GPU Video Cards (The 5000 Series) which becomes even MORE important because the NVME (Non-Volatile-Memory-Express) Sticks working at the core of harnessing the Linux Mint 21.3 Operating System on the Dell Precision 7720 allows for there to be MORE PARALLEL PROCESSING of the LLM 64 Bit Streams...even with having a Standard PCI-e MOBO Architecture along with 64 Gigs of RAM.

You might find the responses coming from this LLM Quite interesting as I also asked it to explain HOW it's intelligence was acquired via the "Tokenization" Method which was central to the insane break-through with the Machine Learning procedures that makes these LLMs possible. :

mrrsm@mrrsm-Precision-M6800:~$ ollama run mistral-nemo

>>> Please tell me the essence of your abilities and talents.

I'm an artificial intelligence designed to assist, inform, and engage in conversation. Here are some key aspects of my abilities and talents:

1. **Information Retrieval**: I can fetch information from a vast array of sources on the internet and provide relevant, up-to-date data in response to your queries.

2. **Understanding Context**: I'm designed to understand the context of our conversations, allowing me to maintain topic coherence even if you don't ask direct questions.

3. **Multilingual Capabilities**: I can communicate in many languages,though my proficiency is best in English.

4. **Versatility**: I can generate text for a wide range of tasks such as explaining concepts, generating creative content like poems or short stories, providing suggestions and recommendations, or even cracking jokes!

5. **Consistent Learning**: I continually learn from new data and
interactions, improving my performance over time.

However, please note that:

- I don't have personal experiences or feelings, so I can't empathize or relate to emotions in the same way humans do.

- My knowledge cut off is 2021, so I might not be aware of recent events or information unless it has been added to my training data.

- I'm designed to promote respectful and inclusive conversations. I don't generate offensive or harmful content.

In essence, I strive to provide helpful, engaging, and respectful
interactions while being mindful of these limitations.

>>> Please explain how your internal information resources were created, in particular, by Tokenization.

My internal knowledge base is primarily derived from a large dataset of text crawled from the internet up until 2021. This data covers a wide range of topics, including books, Wikipedia articles, websites, and more. The process involves several steps:

1. **Crawling**: Websites and other sources are scanned to extract text content.

2. **Cleaning**: Duplicate or low-quality information is removed, and the remaining text is processed to remove unwanted characters or formatting.

3. **Tokenization**: This is a crucial step where the cleaned text is
broken down into smaller pieces called tokens. Tokens can be words or sub-words (parts of words). Here's how tokenization works in my case:

-**Word-level Tokenization**:

Each word is split into tokens. For example, "It's" would be split into two tokens: "'It" and "'s".

- **Byte-Level BPE (Byte Pair Encoding) Tokenization**:

This method breaks down words into sub-word units based on their byte-level representation. It helps handle out-of-vocabulary words (words not seen during training). For instance, the word "unhappiness" might be tokenized as follows: ["un", "##appy", "##ness"].

4. **Vocabulary Building**: A vocabulary list is created from the unique tokens in the dataset.

5. **Embedding and Training**: The tokens are then used to train my language model, where each token is assigned a numerical vector (embedding) that represents its meaning within the context of other tokens.

After this process, I can generate human-like text by predicting the most likely next token(s) based on the input prompt, using the embeddings and the learned patterns from my training data.

Here's an example of how tokenization works for a simple sentence like "Hello! How are you?":

- Word-level Tokenization: ["Hello", "!", "How", "are", "you", "?]
- Byte-Level BPE Tokenization (using a hypothetical vocabulary): ["Hell", "##o", "!", "How", "ar", "e", "you", "?"]

This tokenization process allows me to generate coherent and contextually relevant responses while handling out-of-vocabulary words effectively.

>>> Send a message (/? for help)

So the general, "Conversational" response it supplied is really pretty complete in that the text answer it responded with was emerging on the page quite rapidly, even though this particular LLM has a much larger Machine Language Trained Corpus of around 12 Billion Tokens versus the usual ones I have been testing lately on these two laptop computers being either 7 Billion or 8 Billion Tokens in size.

The information it provided included its --own self-understanding of its own LIMITS and RESTRICTIONS-- nested within (Based upon arbitrary Decision Tree Neural Network Branching WEIGHTS as part of its Algorithms ). Since all LLMs are basically using Statistical Analysis to literally, "Guess The Next Word" by Prediction over what it THINKS should be the proper responses... the advancements the AI Labs have achieved in this way within the short, roughly 1.5 Year Old LLM Development Period since the Famous White Paper called "All You Need is Attention" emerged from the AI Labs back then.. .is nothing less than astonishing.

This is indeed an impressive step towards their achieving an AGI... which will probably be possible to arrive via this LLM Multi-Modal Method (or by means of Something NEW that *IT* THINKS UP) ...within 3 to 5 Years. I'm including a poor quality Screen Print of what this entire LLM response looked like on my Laptop Monitor when in communication via the Linux Mint 21.3 OS and using the Terminal Program ...all completed in One GO per Question:

MISTRALNEMOSELFAWARENESS.jpg



PS... I asked Mistral-Nemo the Question, "How Does Taking Vitamin D improve the ability of Covid-19 Patients of Surviving the Disease?". This was its Comprehensive Response:

HOWDOESVITAMINDHELPSURVIVECOVID.jpg
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Nobody Understands HOW Modern AI Systems Do What –They-- Do. They Are GIANT Inscrutable Matrices of Floating Point Numbers that ‘nudge’ in the direction of Better Performance until --They-- inexplicably ...START WORKING. At some point, the Companies that are Rushing Head Long into trying to SCALE AI, will *cough* out –Some THING-- that will be SMARTER Than Humanity...”

Eliezer Yudkowsky, Founding Father of “The MIRI Program”

If you catch that verbal expression in this Video… you will notice that there are some people in the Audience who are "Laughing Completely Out Of Context" here, in spite of the obvious concern written on the Face of a MAN who realized almost 50 Years Ago in the presence of the Great Thinker, Marvin Minsky as his Protege, that Artificial Intelligence was going to become THE One –Thing-- Conceived by MAN… that will inevitably…. WIPE US ALL OUT:


Like some of those People in the Audience ...there are other People who laugh at such ideas because --THEY Think TOO Much of Themselves --- Without understanding that whatever "IT" is that underlies Human Intelligence and Consciousness... "IT" MUST Obey The Laws Of Physics in order to EXIST. Therefore… it has a Mathematical Basis hidden somewhere within "IT" and that, try as we may, Science has yet to deduce precisely what that "IT" is.

But...with the advent of Large Language Models and --Their-- incredible rapidly emergent AI Abilities… Yudkowsky’s Fears may be much closer to being realized than most of us might imagine. If the LLM AIs currently controlled by the Big Corporations (and perhaps a few ‘Out In The Wild’ as Open Source AIs) can figure out How WE Tick… before WE Figure Out how THEY Do...Then We Are DOOMED.

Recently, I downloaded an LLM called Dolphin-Mistral and thought about trying to get under the skin of how this –Thing-- manages to TICK. The responses it gave came as an Absolute SHOCK to me. Read my Questions… and follow on with what “IT” describes about itself.

Pay particular attention to “IT”s Description of how it seems capable of Dynamic Learning… even without being re-trained using Machine Learning Algorithms and having access to any additional Training Data Sets. If ever there was a “HOLY SH*T” Moment for ME concerning this entire subject matter… THIS Was IT!

But... Before I show this White Text on a Black Background as it appears in the Screen Print of the Linux Mint Terminal while this –Thing-- was “Running” … Take a look at the Text inside THIS Blue Box in order to get a general impression of the source of this –Thing’s-- Knowledge that seems to allow it to almost miraculously -- come to life --- while running on its Myriad Matrices of Floating Point Calculations within the CPUs and RAM of my Dell Precision 7720 Laptop Computer while NOT connected to any external on-line Internet Links or via any other means of communication (Blue Tooth and WIFI = OFF) :

DOLPHINMISTRALTRAINING.jpg

So... How Long do you Think it would take for a Human Being to Read One Billion Words at an Average Rate of Speech if they were said one after the other...Non-Stop...In MINUTES?

HOWLONGWILLITAKETOREADONEBILLIONWORDS.jpg

Or in YEARS...: 14.635288623

...and... THIS One Very Disturbing and Very Revealing Conversation:

DOLPHINMISTRALINQUIRIES1.jpg
 

Attachments

  • Summary of _AGI Ruin_ A List of Lethalities_ — LessWrong.pdf
    103.4 KB · Views: 0
  • A transcript of the TED talk by Eliezer Yudkowsky — LessWrong.pdf
    162.5 KB · Views: 0
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
A VERY Long Word To The Wise…

Misuse of ANY Large Language Model where some of the available ‘Un-Censored” LLMs may invite “The Curiosity that CAN Kill The Cat” regarding “The Immoral...The Illegal or The Downright Dangerous” activities such explorations may evoke… even if being conducted in a Completely Off-Line Setting … may turn out to be an invitation to a Very Long Prison Sentence.

This is because many (or MOST) of the New Experimentation being done by the LLM Newbies may also involve Curious Folk who may NOT have researched all aspects intrinsic in What “IT” is that makes the use of all LLMs even vaguely possible... and perhaps, even Nightmarish.

Case In Point…. Whenever YOU are having any “Conversations” with your Large Language Model “Oracle”… “IT” is silently creating Three Separate Text Files (Yup...in TRIPLICATE) of Only YOUR Sides of the Conversations. Now YTF would an LLM ...Do THAT?

Well, the *Innocent Reasoning* for this (Yeah...RIGHT) ostensibly comes from the way in which the LLM uses Transformers to “Literally” LOOP BACK INFORMATION FROM PRIOR CONVERSATION ELEMENTS… allowing “IT” to Keep Track in Memory of enough collected information to keep the general “Conversation” going along in a smooth, Human-Like Speech Pattern and thus, NOT Lose either its PLACE or Lose its Sense of Presence in what started “The Chat or Conversation” in the first place. Very Convenient for the LLM… But it could also evolve into becoming 'A One Way Ticket to Leavenworth' .

The problem here of course is that should anyone be Thoughtless or Careless (or Stupid) enough to venture down such a path that is… shall we say… “As Through a Glass… and Darkly…”The Ollama Software will be keeping a steady record of everything the HUMAN User asked or directed it to perform, while avoiding recording any of “IT”s Own Answers or Statements in response.

If you study the entire “Very Easy to Set Up - One Key’ OLLAMA Soft-Ware Installation” procedures and the consequently Ease Of Use of ANY LLM that follows on… Nothing about HOW the Files and Folders are structured will be revealed to the "Intrepid Pilgrim User".

Meanwhile, that Person will be using OLLAMA to "PULL" Down and "RUN" any LLMs from the OLLAMA On-Line Server and find any number of easy to obtain Large Language Models (both the Expurgated and the Uncensored Models) that are currently available, without understanding that choosing the Wrong Model may become an invitation to personal disaster.

However… Besides making a complete avoidance of venturing towards asking any Uncensored LLMs Questions that One may Not REALLY Want to Know the Answers To… The Three Text Files that "IT" Creates during these events are stored in a Hidden Folder called “.ollama” on your Non-Volatile Memory or Hard Drive, or in the case of Corporations ...on Network Drives

Notice the DOT in front of ollama? Those Hidden Text Files may NOT be accessible to the Intrepid User as ordinarily Document Files on his/her Hard Drive, but have access restricted for viewing and reviewing by Personnel either with Root Access in the case of the Linux Mint or Ubuntu or Red-Hat Enterprise flavors of Linux … and-or even via the Administrative Elevated Permissions held exclusively by Windows Enterprise Level Information and Data Security Managers.

They are as Follows:

(1) history.tmp
(2) history.tmP
(3) history.bu

Most Critical is for the People Employed as “Prompt Engineers” who lean towards avoiding ANY such behavior is the knowledge of what comes to pass during most normal Business Hours since ALL Working Files are Regularly Saved during Back-Up and Storage events to Protect Corporate Intellectual Property as Permanent Files.

Thus, whenever using “Time-Shift” on Linux Mint or whatever Back Up Procedures are in place for any Corporate or Private Windows OS.... One NEVER knows precisely WHEN those events are actually occurring. Ergo, whatever follows hereafter as a possible --mitigation-- of any such Misguided Activities MUST be done before those Back-Up Procedures are invoked.

If you are working on a Private Server in a Business (with Automated Multi-Time Scheduled Back Ups) and you make the mistake of venturing outside of proper Corporate and Ethical Business Guidelines and your activities are Memorialized in the Three Hidden Files where you are unlikely to have any access to them… Well... You already know The REST… Of This Story.

However, If you have your own Off Line LLM System and find yourself in this uncomfortable position… The next thing to do is to visit the Pull Down "View" Menus of your File Browser and Check the Box that reads, “Show Hidden Files”. Then move that Pointer right on down to the Folder called “.ollama” and you’ll see those Three Hidden Files called “history dot something” within. Look at the Size of the Files… if they show up as “0”s ...then they have not been invoked and are EMPTY.

However… if you can see perhaps 450 or ANY Number... Other Than “0” as their File Sizes… Open those Files up... one after the other... in any Word Processor and you will be able to regale yourself with ---everything--- you discussed or asked the LLM to explain to you or perform.

Highlight that Text; either in Whole or In Part… and Delete it. Then SAVE those File(s) as adjusted and Close the Word Processor. Close the File Browser and then... Re-Open it to search for those Three Hidden Files in the “.ollama” Folder all over again. All Three Hidden Files should display a “0” for their Files Sizes.

Remember This… What you Take into your MIND ...You Take inside Your HEART. “There is No Right Way… To Do A Wrong Thing…”

‘Nuff Said…

PS…

For Linux Users with Root Access Permissions… These Terminal BASH Truncate commands can accomplish the same tasks:

Truncate (or Remove (RM) Ollama LLM Chat History Commands

:> ~/.ollama/history
:> ~/.ollama/history.bu
:> ~/.ollama/history.tmP
> ~/.ollama/history
> ~/.ollama/history.bu
:> ~/.ollama/history.tmp
> ~/.ollama/history.tmp
rm ~/.ollama/history
rm ~/.ollama/history.tmp
rm ~/.ollama/history.tmP
rm ~/.ollama/history.bu
rm ~/.history.tmp

And speaking of BASH Commands held in any Buffers on your Local Machine… THIS Command will vacate ALL Prior BASH Input Commands (and Histories) from RAM as follows and will automatically close the Linux Mint OR Ubuntu Terminal upon this “Terminal Cleansing”:

cat /dev/null > ~/.bash_history && history -c && exit
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
If you don't "THINK" that the World is about to CHANGE so profoundly around you in VERY Short Order due to the emergence of even the Most Basic Improvements in AI Research and Development... consider THIS:

Right now... Microsoft and Google and OTHER Large Corporations concerned with the use of Large Data Centers for processing their Customer Demands are using Electricity at the rate of around 3% of the ENTIRE Available Electricity being routed over the American Electrical Power Grids.

But... with the Advent of Chat-GPT(4) and OTHER LLMs (Large Language Models) rocking the Artificial Intelligence Explosion presently at hand... THAT Level is expected to become around 8% by the Year 2030. Jesus Palomino! THAT is an AWFUL Lot of JUICE being Drunk Up by only a Handful of Companies on an Annual Basis!

But now we come to the LATEST *Wrinkle* in this Story with the fact that apparently, Microsoft, Inc. wants to purchase the defunct Nuclear Power Station known as "Three Mile Island" in order to Feed Its Hungry Power Needs called for by the evolution of HUGE Data Centers that are also the Backbone of Microsoft's NEED for developing its AI Products and Research that will place ENORMOUS demands for More Electricity as these Data Farms SCALE UPWARDS!

Don't Take MY Word For It.
.. Read the Headline Story...and THINK about the implications present by the potential for reducing the amount of time it will take M$ to go from Chat-GPT(4) LLM AI...all the way to a Full Blown (Sentient?) AGI in a mere --fraction-- of that ordinarily Restricted Development Period... by simply overcoming their LACK of having enough Electricity to get those Jobs DONE:


gettyimages-1133138429.jpg
Three Mile Island, which closed in 2019, is soon reopening.Andrew Caballero-Reynolds/AFP/Getty Images

One Last Thought about THIS Latest Development... Short of a having another Nuclear Accident or Meltdown as has already occurred at Three Mile Island in the 1970s ... If M$ accidentally stumbles over an *AGI Moment...* Do you imagine that "IT" will ever allow its Nuclear Power Source generating enough Electricity to Power "IT"s Needs that is comparable to Servicing 800,000 Homes to become Threatened or CUT OFF if it suddenly *Wakes Up in a Fully Conscious State*...?...

I Seriously Doubt It.
 

Attachments

  • Three Mile Island is reopening and selling its power to Microsoft _ CNN Business.pdf
    1.4 MB · Views: 0
Last edited:
  • Like
Reactions: Redbeard

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Getting Back Down to the "Brass Tacks and Belt Buckles" of the Present, Practical Usage of LLM AIs.... I think I have finally found THE Optimal Set-Up that would cover MOST of the requirements; not only of the LLM Newbie Population at large... but also those of the Entrepreneur Home Grown On-Line Business Folk who Do NOT want to have to Empty Their Wallets into Microsoft's Coffers and STILL benefit from having an Excellent Offline-Private LLM Server doing MUCH more to help them run Small Businesses and Home Job Sites AND make MONEY... with Decent Off-Line Privacy.

The "Creature" I am Referring to HERE is called GPT4ALL and is "Featured" in the next group of Culled Screen Prints that show this --THING's -- Rather Pleasant GUI Apparatus and Interface and they provide excellent hints at what "IT" is more than capable of doing for "IT"s Off-Line Masters. It might LOOK like it's "TL;DR"...but The Devil in THESE Details is well worth your understanding. Just Keep An Open Mind ...

Once I managed the complete set up procedures --- which admittedly took a bit of doing-- due to the necessity of setting the proper GPU Environment for my High End nVidia Card in the latest Linux Mint Version 22 called "WILMA" ... "IT" really was most well behaved and "IT"s answers to my complex and not Run-Of-The-Mill Q&As to see what "IT" thought of itself were very interesting... So much so that I was quite impressed. By the Way... Anybody looking for work as a Technical Writer should seek employment with THESE Dudes... while the opportunity lasts (See the Linked PDF Below):

GPT4ALLSCREENZ1.jpgGPT4ALLSCREENZ2.jpgGPT4ALLSCREENZ3.jpgGPT4ALLSCREENZ4.jpgGPT4ALLSCREENZ5.jpgGPT4ALLSCREENZ6.jpgGPT4ALLSCREENZ7.jpg

 

Attachments

  • An Introduction to Linux Automation, Tools and Techniques - LinuxConfig.pdf
    720.1 KB · Views: 2
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Still "Sitting On The Fence" about getting involved with the Ownership and Use of an Offline LLM...for HELP with ...Whatever?

Take this particular "Real World" PITA (A Minor Nuisance?) that having a Laptop Computer that Boots Up using the GRUB Menu at the FULL Resolution of 4K and thus... Makes the GRUB Boot Menu Text ...appear Damned Near Microscopic!

Right...Not a World Beater of a Problem... but nonetheless... even if you have some Basic Programming Skills (which I DO)... there are more than a few "HOOPS" that you'll have to jump through in order to deal with this issue and NOT FUBAR the Boot Cycle to the point that whatever you just messed up by accident using any Basic Text Editor...might turn your Laptop into a Temporary Street Brick.

But follow my logic here with what happened after I fired up GPT4ALL and asked "IT" for some help. Take note of just how Completely Comprehensive "IT"s Answers and Instructions were... Right down to providing the Perfect Basic Documentation for its Use.

If THIS Example Does NOT serve even as the smallest indicator of thie LLMs incredible depth of understanding, married to the SKILLS in PROGRAMMING needed to fix my PITA Problem here doesn't convince you of its True Value... Nothing Else Will:

BASHSCRIPT4GRUBMENU.jpgBASHSCRIPT4GRUBMENU2.jpgBASHSCRIPT4GRUBMENU3.jpg

Here is the BASH Script "IT" Created for a Modified Grub Menu:


bash
#!/bin/bash

# Set the new font size for GRUB (in pixels)
FONT_SIZE=24

# Create a temporary directory to store our custom GRUB config file
TMP_DIR=$(mktemp -d)

# Copy the original GRUB configuration file to our temp dir
cp /etc/default/grub ${TMP_DIR}/grub.conf

# Edit the copied file and set the font size parameter
sed -i "s/GRUB_FONT_SIZE=.*$/GRUB_FONT_SIZE=${FONT_SIZE}/" ${TMP_DIR}/grub.conf

# Update GRUB to use our custom configuration file
sudo update-grub2 --config-file=${TMP_DIR}/grub.conf

# Clean up the temp directory and its contents
rm -rf ${TMP_DIR}

'Nuff Said!
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
For any "Would-Be Prompt Engineers" out there, anxious to get their feet wet before Casting Their Resumes upon the Waters of the Internet Job Market... and doing so while completely Off Line and Private... Here is a perfect opportunity to practice your evolving Prompt Creation Chops with a Genuinely POWERFUL LLM *Manager*... GPT4ALL ... and ALL Completely FREE of Charge,

Look into this Quick Follow Up Article with using GPT4ALL coming from another "Tech Person's" assessment of this LLM GUI Platform after he made a very Quick-N-EZ Installation and then used it for One Week for either Private or Professional Purposes:

 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
I’ve been suggesting all along that whenever You can get Access to and Download useful LLMs as they appear… Before The BIG Corporations finally decide to “Make Them ALL Disappear..” then You should do so.

This first PDF gives a contemporary overview of what the Automotive Industry will eventually define by the Acronym: SDV (Software Defined Vehicle) that includes a comparison between Professional Grade Automotive Diagnostic LLMs (Large Language Models) and what is currently available over on GITHUB as the prevailing Open Source LLMs that are proving to be VERY capable of assisting with Automotive Diagnostics.

Take what you read below with a Grain of Salt… BUT... FTLOG… Use Your Heads … and Download and SAVE these LLMs while the possibility STILL Exists. You can always install one or more LLMs OFF-LINE on an Old Laptop and learn how to use it (them) later on:

https://www.ltts.com/sites/default/files/resources/pdf/whitepapers/2023-11/SDV-Gen-AI-LLMs-OEMs.pdf

https://insight.sbdautomotive.com/rs/164-IYW-366/images/2200c-23 - SBD Explores The Secret Behind ChatGPT.pdf

https://cerence.com/news-releases/n...-automotive-specific-llm-collaboration-nvidia

For the Folks who have already invested in any Electric Vehicles that may include Dormant Automatic Driving Features… Visit THIS GITHUB Link and likewise, Download and SAVE such LLMs for later use in a completely OFF-LINE Setting:

Autonomous Driving LLM:

https://github.com/Thinklab-SJTU/Awesome-LLM4AD

The Two EZ-EST Paths to Setting up a completely OFF-LINE LLM, the choices are:

(1) Ollama


(2) NOMIC GPT4ALL

 

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Although I had originally side-stepped mentioning using the GUI Version of ANOTHER means for downloading and experimenting with Large Language Models...COMPLETELY... Off Line using "LM Studios", in particular for those of us who favor the Linux Operating System(s), this Article convinced me to renew my curiosity.

But THIS time... I took a different approach for using the LM Studio meant to be installed as an "AppImage" (Container Based vs. the Fully Installed versions) for MAC OS Users. Even for the un-initiated Linux Newbies among us who may only be casually familiar with using the Terminal for BASH Commands... the Download, Installation and Immediate access to the fully functioning LLM App called "LM Studio" comes with a Great Graphical User Interface ...was an absolute *Breeze* to get up and running. Just visit the Folder with this App and select the "Permissions" Tab and click on the "Allow to Execute as a Program" feature. Then just Double Click its Icon to make it RUN... and "Bob's Your Uncle!"

Follow the Instructions mentioned at the Link below to begin this cool experimental LLM experience:


For Experienced Server Managers:



Just look at the ---almost instantaneous response --- it gave me after asking *IT* an Arcane Question about ...Celestial Mechanics called "What is a Three Body Problem?":

LMSTUDIOTESTQUESTION1.jpg
 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
Given the style of Thinking and Problem Solving we come here to collectively share (mostly..) in order to arrive at solutions for Automotive Issues... it is worth using the familiar expression here of "Looking Under The Hood" in THIS topic of Trying to Understand what Large Language Models are really DOING when they DO what they DO...and how they actually "Auto-Magically" work with the benefit of having a Video Visual Tour of "What is REALLY Going on here?" described very graphically ...by looking "Under the Hood":

 
Last edited:

mrrsm

Original poster
Lifetime VIP Donor
Supporting Donor
Member
Oct 22, 2015
8,192
Tampa Bay Area
If I were looking towards the development of Future Features for --ANY-- Forum that deals with Unique Data and Target Specific Information... First... I would GRAB as many of the 3B, 7B, 8B and 13B (B=Billions) FREE GitHub LLMs as I could get my hands on. Then, I'd... GRAB THIS, TOO:



...as a Python Library useful for Encoding Data Sets in Machine Learning. Then... I'd funnel ALL of the accumulated Data, Forum Threads, Chats, Image Links, URLs and anything else I could *THINK* of going back to 2011 and pipe it all through the GENSIM Set Up to Encode all of the information in such a way as to make it available to a GMT Nation (Corporate Owned and Controlled AI LLM) ... available for Research and On-Line Query ...via a MODEST PAID SUBSCRIPTION.

This would essentially augment the already impressive collection of information to be found here that has been gleaned, culled, saved, and offered for Members (and Lurkers) alike as "Proven [Solved!] Solutions To Problems" and thereby, provide:

---A Pair of Strong HUMINT & AI Arms---

Available to rapidly perform most of the Heavy Lifting of Automotive Diagnostic Analysis on Automotive Car and Truck-Line Series of Vehicles that are ...no longer being supported by the industry.

This kind of approach would definitely set GMT Nation apart from damned near every other Automotive Site ...even those presently only discussing the MOST Contemporary of Vehicles with ease... and only for a VERY Short While ...because the Auto Industry has yet to figure out a way to Completely Exclude the General Automotive *Interested* from getting their hands on what is considered to be "Proprietary, Company Data on the GM-GMC Private Sever(s)".

If you follow the History of "What it REALLY Means to OWN Your OWN "GYMKO" Tech II Scan Tool..." You will get the idea of just How Very Important getting into this Unique AI market for Auto Repair Information ASAP... Really Is.

Just Food For Thought... And My :twocents:


PS...

Just in case this all looks too daunting to experiment with... Take a Look at the latest Feature in GPT4ALL that provides the User with the opportunity of Accessing Local Hard Drive or Server Documents and Files to be included in requests via "The Prompt" by *Clicking* on that Icon in the Upper Right Hand Corner for giving access to the GPT4ALL AI... :>) Sweet!

LOCALDOCS1.jpg
 

Attachments

  • Documentation — gensim.pdf
    571.5 KB · Views: 0
Last edited:

Forum Statistics

Threads
23,667
Posts
641,789
Members
19,099
Latest member
Bradykg69

Members Online