Artificial Intelligence: BingChat (Copilot), ChatGPT, etc
Suggestions for safe AI development:
- Stop having AIs imitate people in the sense of having an ego or self-awareness.
- AIs should not be given the power to run things beyond local machine control.
- Deal with the copyright issues, etc.
- Increase penalties for the criminal misuse of AI technologies.
- Law enforcement should become proactive about the criminal misuse of AI.
LLMs are useful for gathering and synthesizing public information and for problem
solving. Build into them whatever safeguards are necessary.
Appendum: As we are now in the agentic age, some limited control
is being given to agents to get work done, but care should be taken fto prevent
possible harmful side-effects.
Hyperlinked Categories:
- Jump to section Academia (the good and the bad)
- Jump to Agentic Internet and Architectures The future is Agentic -- on the desktop and on the internet.
- Jump to section AI-and-LLMs
- Jump to section US Government
- Jump to section Archaeology and Old History
- Jump to section Avionics and Flight
- Jump to section Baby Boomers
- Jump to section Biology-Botany
- Jump to section General Business Information IT, security, general information
- Jump to section Chemistry and Biology
- Jump to section Chess
- Jump to section Christianity, Religion, and Esotericism
- Jump to section Computers (Includes cyber security)
- Jump to section Conspiracy Theories Flat earth.
- Jump to section Cooking
- Jump to section Economics and Finance
- Jump to section Education, Academia, Universities
- Jump to section Entertainment
- Jump to section Ecology, Recycling, and green chemistry
- Jump to section on Every-Day Stuff
- Jump to section on English Grammar
- Jump to section Health & Medical, Mental Health, and Accident Prevention
Includes pet care.
- Jump to section History WWII, etc.
- Jump to section LLM Humor
- Jump to Mathematics Page
- Jump to section Mathematics Education
- Jump to section Miscellaneous (hardware and software)
- Jump to section Nature's Mysteries
- Jump to section On-Line Safety
- Jump to section Philosophy and Cognitive Psychology
- Jump to Physics & Astronomy Page
- Jump to section Programming (Python, LaTeX, Mathematica, theory, etc)
- Jump to section Science Development
- Jump to section Social Media
- Jump to section Technology (cable modems, LANs, WANs, space technology)
- Jump to section Weather and Atmospherics
- Jump to section Navigating the work world
- Jump to section Writing Fiction & non-fiction writing, Sci-fi analysis
Note: When I first started to review the reasoning capabilities of LLMs, I spent much time trying to evaluate if they are capable of 'real reasoning' on the level of a human being. After a while I became convinced that they are capable of human-level reasoning, so I stopped commenting on this fact. (When I'm not being precise, I may slip into the mode of interchanging 'thinking' with 'reasoning'. Technically, this is a mistake. Thinking is everything, both physical and abstract, that goes into reasoning, whether of a biological entity or of a machine. Reasoning is the ability to construct logical arguments that can be written down on paper and examined logically.)
Academia
The Case of the Naughty AI. How does the case of Biosphere 2 figure into all this?
Agentic Internet
How does it compare to or complement Shopify and Cloudflare?
AI-and-LLMs
I'm still fairly ignorant about how models are developed, and it shows in my chat with Copilot. Still, by the end of the chat, Copilot and I arrive at a list of ten Laws of Model Development (mostly contributed by Copilot, of course). We also discuss how chaos theory and random matrices will play a role in the theory of model building.
An early version of an LLM assisting the CIA.
Some ideas on how to deal with this.
What to acknowledge and what to let go.
Copilot serves up its own research program to align with Fredric
Schuller's Conjecture.This link is joined with the previous link.
The Turing Test, while historically significant, may have set a questionable precedent for human-AI interaction. Rather than focusing on whether an AI can "pass" as human, the real value lies in how well it provides accurate, useful, and ethical responses. Impersonating emotions can blur lines that don't need to be blurred. --- Copilot [1 June 2025]
Copilot and I haggle over whether it has a personality, value judgments, and a virtual subconsciousness,
that all derive out of the stochastic manner in which it gains information, or rather, "experiences the
World of Information." Copilot seems to have a way to determine what is "best," though I could not get
it to tell how it does so.
Is a chatbot sentient or conscious or something else?
What correlations might be revealed by AI comparing large databases of disparate types?
LLMs are strong where I am weak.
In this test of LLMs, I test the ability of Copilot and ChatGPT-4o to reason with apparently contradictory
statements that perhaps can be made compatible by subtle reasoning. Are they up to it? Let's see.
Which is better for running AI programs and why?
What ChatGPT-4o thinks about my suggestions that 1) there is really only one 'kind' of intelligence
and 2) that intuition may be its own form of intelligence, though science keeps it in the shadows.
I anticipated that ChatGPT would fail to grasp the play on words, but it didn't.
How the chatbots interpret the so-called 'analogy problem' that are actually proportion algebra problems.
BingChat and ChatGPT attempt to evaluate symbolic logic expressions.
BingChat and ChatGPT wrestle with distinguishing between thinking vs reasoning.
The makers of chatbots and AI graphics tools are under legal pressure these days.
It's an uncertain future facing us who love these chatbots.
ChatGPT creatively finds an analogy on demand. Bravo!
Although it succeeded, I had a different analogy in mind. Nevertheless, the bot is
very impressive in its reasoning ability! Could the average human do as well?
for algebra word problems and then disagrees with me that it is being inventive in the formulation
of its list of heuristics. Question: Can a piece of software be creative or inventive? We delve into
this philosophical and practical question.
Does ChatGPT Understand Logic?
I asked BingChat/Copilot to decide
the matter. BingChat claims the LLMs are not good when dealing with
counterfactuals.
Ten AGI Benchmarks.
ChatGPT and BingChat advise me on benchmarks
for the attainment of AGI (Artificial General Intelligence). I suggest that music,
creative writing, and humor may be too much to ask of AGI, and to leave them
as qualifications for Super-AGI (SAGI).
Is ChatGPT capable of grasping humor, even if it can't experience it emotionally? Maybe. Let's see.
I asked ChatGPT to go one level up on the old 'Is it half full or half empty?' conundrum. Perhaps this
conundrum is a useful test for LLMs generally.
in the development of the virtual machine. No Internet search string could return a reply so
well-crafted as ChatGPT gave me. Such a reply leads me to believe that LLMs can at least
understand complex sentences and reply appropriately. Judge for yourself.
US Government/ Gov. Agencies
And the Navajo Reservation.
Archaeology and Old History
Avionics and Flight
Biology-Botany
#Business Applications
Copilot helps me to understand the great benefits for large businesses to use OpenAI
Enterprise. Some applications of this technology will not be a threat to anyone's job
security, as it will be able to do things that no human being could ever anyway, such
as know the company's entire working documentation. Also, it seems that OpenAI
Enterprise's preference to support Python will make Python an even greater lingua
franca than it already is, and help to eliminate the dreaded memory- management
errors of C and C++.
SOC 2 (System and Organization Controls 2) is a security and privacy standard invented
by the American Institute of Certified Public Accountants (AICPA). Trained CPAs perform
the SOC 2 audits, but who checks the checkers for accuracy?
Limited Liability Company purpose.
It's definitions and dangers.
#BabyBoomers
#Computers
The Year of Linux Desktop From Windows to Linux to AmigaOS to maybe LeelaOS?
The Secret AI Code How should we respond to AIs inventing their own secret programming languages?
Re-install Windows11 There's a lot to consider: Original
Product Key, dual boot, wiping the drive first, etc.
What is a use case?
Copilot helps me understand what a 'use case' is
and what the 'case' in 'use case' means. I leave it to the experts to
confirm or deny the conculsions we reached.
What is a 'proof assistant'?
BingChat gave me some good information about
what a 'proof assistant is. They have help to prove some major theorems over
the last 40 years.
Soft Skill needed for Tech
ChatGPT presents its advice on gaining soft skills
needed for the tech of the future.
Is SecureBoot a threat to Linux?
Some Linux users think this is so, while
BingChat (Copilate) does not.
#Programming (Python, LaTeX, Mathematica, theory, etc)
Rust keeps the better parts of C++, while it levels up to Dependent Type Theory.
It's more than mere memory management.
Controversy about using Typescript.
What do they mean?
Copilot weighs in.
Certain performance enhancements are built upon C code, but is it memory-safe?
If it's not broken, don't fix it?
What does it mean?
ChatGPT wrote a Python program for me to find the GCD of two numbers, using the Euclidean
algorithm. Included in this article is a comparison of various other Python programs that do the
same thing, but are coded differently. Very instructive.
How BingChat is able to help me code some simple and some difficult LaTeX encodings.
How Coplot thinks that the C programming language can be made more memory safe
by programmers adopting best practices.
#Miscellaneous
ChatGPT answers my general questions concerning the considerable changes to the US military
ranking systems since WWII. I think it's quite clear that ChatGPT understands the questions posed
to it and presents clear and insightful replies to them. I score 5 points out of 5 to it.
Is Jurassic Park a 'rip-off' of Fantasy Island?
It's difficult to make objective
comparisons on this.
The LLMs read me the riot act for my hyperbole!
Bing Chat and the A-10 Warthog Jet.
#Technology
Should new NGOs be started to oversee/monitor software vulnerability patch rollouts?
Python, datasets, translation of ancient languages, NaNoWriMo's collapse, formal education verses generative AI.
Use your phone instead!
#On-Line Safety
#ConspiracyTheories
Flat-earth idiocy.
What does it really mean? Is it often misapplied? Or is it misunderstanding the flight from negativism?
Be careful out there!
#Nature's Mysteries
My chat with Copilot on the uses and mysteries of boron, and the need to recycle it.
#Science Development
Analogy in Teaching and Scientific Discovery.
Of Planets and Fish.
--- the oddities of some scientific classification schemes.
#Mathematics Education
It comes up a lot in math competition problems and entrance examination problems.
#Chemistry and Biology
Should it be changed or modified or left alone? At the very least, we should better understand the system in use.
Veritasium looks at the Haber-Bosch process that helped make the modern world.
Teaching acid-base ionizations
from the viewpoint of conserved quantities, thereby turning
it into an algebra word problem.
Copilot explains to me the process to derive aluminum from bauxite by the Bayer process for alumina
extraction and the Hall-Heroult process for aluminum production. The recycling of aluminum is
also discussed and rated as preferred ecologically.
Both ChatGPT and BingChat attempt to balance chemical equations. Both are able to get a lot of the
demonstrations correct (at least in form), yet neither are consistent at producing accurate results.
Hopefully, they will both improve in time.
ChatGPT reveals how proficient it is as a chemistry professor when it comments on my
copper plating experiment from over 50 years ago. I'm not a chemist, so I need a real
chemist to read it to evaluate it.
Zen and the Art of Motorcycle Maintenence
Philosophy and Cognitive Psychology
My claim is that it is necessary to hold to this principle in order to avoid contradiction in one's thinking.
I began my inquiry being interested in sources of random and near-random processes or signals for their
possible practical means to generate random events for computer modeling or random-number generators.
But the discussion ended up in the philosophical, in the middle ground between determinism and randomness.
And finally, we ask the probing question if wind chimes can be used to generate a random number sequence.
The Problem of Determination vs Freewill with AI. John holds to two conflicting beliefs. Can the
chatbots figure out how badly John is conflicted?
Bayesian priors and educated intuition.
Economics and Finance
ChatGPT 3.5 tries to explain to me the complicated world of monetary exchange rates
and arbitrage.
I could use a report
that tells me the agriculture, economics, industries, and geography of a region. It should include:
1) population (and how it's trending), 2) GDP (if available), 3) natural resources, and
4) Economic Outlook Index. Who wants to champion this cause?
Education, Academia, Universities
Have you ever heard of this university of higher education? Maybe it should rebrand itself for the 21st century.
Some tips on good problem solving.
How academia is more dependent on the humble scratch paper than it gets credit for.
Entertainment
Two Hollywood movies, two songs, two countiries of Greece and Austria, followed by
some rules to live by and wisdom from Robert Pirsig.
Dr. Arroway (Jodie Foster) kept saying she needed a "primer" to decode the Vegan message.
Negative Gaslighting from Actors and Directors.
A comparison theme: a group of has-beens that make a comeback after a daring
and dangerous voluntary mission.
Fiction and Non-fiction writing, Sci-fi analysis
Simply told, they are "true stories, well told." (Gutkind)
Examining and re-examining the term in the modern world.
--- existential crisis, dissociative fugue: the search for identity and meaning.
Navigating the work world
Why it's so hard to reform itself.
Ecology Recycling, and green chemistry.
Spare the Nitrate and Spoil the City.
Iowa's love-hate relationship with nitrate fertilizer
and its runoff problems.
The Challenges of Municipal Recycling.
A Better Job at Recycling Aluminum.
Norway's Hydro company is showing us a better way to do this.
My chat with Copilot
about sequestering CO2 from burning waste by photosynthesis.
Various uses of artificial sand.
My chat with Copilot on recycling resources
AI used to identify potentally toxic chemicals.
Green chemistry.
Recycling Old Medications.
Keeping the water supply clean.
My chat with Copilot on the uses and mysteries of boron, and the need to recycle it.
Weather and Atmospherics
Why are most UV alerts in the afternoon rather than in the morning?
English Grammar
An arcane English grammar rule.
English grammar can be subtle, even to those
who have studied it a long time.
seems to lie
one layer beneath the obvious. A bit of 'Einsteinian"
thinking comes to the rescue.
Health & Medical, Mental Health, and Accident Prevention
I approach the question from epistemology, while Copilot from psychiatry.
The Pitt vs your average hospital ERD. Extra: How wars benefit surgery and other medical procedures.
The mathematics of how the body removes the drugs that get into it. (SymPy graphic included.)
If you're watching The Pitt or some other TV medical program, this might be useful.
Which specie gets the theobromine?
The shelf or the frig?
Means of transmission of avian flu to humans and some means to help prevent that happening.
This can happen in water near a dock. Is it mysterious or even real?
Some people I have known for a long time have had severe personality changes that I cannot understand.
Should someone tell them?
More to the story.
In pets too.
Supplements for brain health.
This is the opposite of the placebo effect.
Takeout food (and other foods) that contains rice may cause you trouble if not refrigerated properly
and reheated properly before eating again.
Lyme Disease is a multifaceted serious disease caused by the bite of a tick.
A claimed health benefit, but which product to purchase?
Lectin can be safe or not, depending on how the food is prepared (mostly cooking) and by the
sensitivites of the consumers. Some lectins are called "antinutrients." There is a lot of nutritional
information in this article.
--- not just the mind!
--- It can happen after just a cold or flu!
Cooking
To use or not to use?
Every Day Stuff
Have you ever tried to define the term?
History
I suppose that Hitler and the Japanese militarists started their wars based on what can only be called "magical thinking."
In particular, WWI and WWII: a 'good' war movie, what is an antiwar war movie?.
Post-war prosection? Why or why not? The Bigger Picture.
What went wrong? There were good reasons to go ahead with Market Garden.
Mostly, I suppose, but what about the so-called Blitz over England?
If Plato was sincere about Atlantis, there are a few unanswered questions that need answers.
Why were they so much less prevalent than those of the Plains Indians and the Indian uprisings of Arizona?
We investigate the claim for the Plains Indians and the Pueblo Indians of Arizona.
Copilot wants us to know.
LLM Humor
after hearing of my complaint that coffee makers are sometimes over-engineered.
Chess
Christianity, Religion, and Esotericism
.
From Babel To Revelation's Babylon. How Babylon, the Persian Empire, and Susa (ancient Iran) interacted.
The different leadership styles of King Saul verses David. What did God have to say about them?
Why did the King of Sodom offer gifts to Abraham?
Between the First and Second Temple periods, the written Hebrew language changed,
largely due to Aramaic influence
from the Neo-Assyrian, Persian, and Babylonians
governments and cultures.
Why did Samaria stand out as an oddball province in Israel during the time of Jesus?
Neo-Assyrian, Babylonians, Jewish Hasmonean rulers, Jesus and the woman at the well."
An example of science in the Bible, and why the unbelievers can miss God, though He is
ever-present to each of us. (How much do we really want to connect with God?)
I asked ChatGPT 3.5 and Copilot to speculate on the possible mysterious physical origin of the
Christ. I call this speculation the "Dark-Horse Hypothesis of Christ."
Other content: esoterica, Melchizedek.
I define the term 'Hittiteism' to mean "a proposed proof that some aspect of the Bible can't be correct,
because of some alleged undeniable historic, scientific, and/or philosophic argument."
I found its answer remarkably scriptural. Still, the ethics of 'company loyalty' is a gray
area to me.
It was helpful, though I end up with no definitive answer.
I think that this time we came up with the definitive answer, at least for myself.
Some say Paul, some say Luke. No attribution was on the document.
The subtle parallel of Christ as sacrificial lamb, yet risen, and of Antichrist's mortal head wound, yet he survived, is explored.
According to Genesis 2:7, God made the body of the First Adam out of the earth, with no earthly father involved. And Luke called the First Adam the Son of God (Luke 3:38). Jesus, the Ultimate Son of God, and the Last Adam, was made (bodily) of earth in the womb of his mother Mary (as the body of Mary was ultimately a product of earthly materials), and with no human father. So, the naysayers against the Bible may well say that Christians teach the 'myth' of the virign conception of Christ as an unproved miracle, but that belittles or ignores the intellectual reality that this sharp parallel between the 'conception' of Adam and the conception of Christ was mysteriously arranged by such disparate Bible figures as Moses (Genesis), Luke (Book of Luke), and Paul (1 Corinthians). So, how were they able to collude to produce this fascinating (and unobvious) Bible parallel?