Is MS Copilot a search engine?

Status
Not open for further replies.
Allow me some time to respond to the bulk of your post.

I'll stipulate to your description of human "knowledge" and "belief", for now.
 
Last edited:
I never asked it what the square root of 4 is.
No. You asked it a bunch of different questions - questions which you don't know the answer to, and therefore are unable to determine the truth or falsehood, or even logical validity of.

So don't quote chatbots, since they cannot be trusted as a source of truth.
 
Write4U:

Here's some useful advice: when you find yourself in a hole you created, stop digging.

You're making yourself look even more foolish than usual, with your pathetic attempts to defend your claim that MS Copilot is a search engine. And all this just to pretend that you weren't ignorant about the difference between a search engine and a large language model AI in the first place, which clearly you were.

Really, who do you think you're fooling? Why continually double down on the stupid?

My guess is that you actually have nothing better to do with your time, so it's just a matter of posting any old thing, for you, as long as it keeps the conversation going. Am I right?

Does it really not matter to you if your readers' perception of you is that you sound like a senile old fool who is pretending to know and understand stuff he is too proud to ask about?

Do you just want company? Is any interaction better than loneliness, even if the feedback you get is mostly negative due to your pretences and your unwillingness (or inability) to learn anything new?

I'd love to know.

Anyway, back to your latest string of inanities.
But that addresses only subjective (anthropomorphized) knowledge, i.e. "belief". Beliefs are human qualities.
Sarkus told you that Copilot doesn't know anything. That is, it has no store of "facts" it can draw on, or anything like that.

Here you are, agreeing that it is no beliefs. But knowledge is typically defined as justified, true, beliefs. So, you're arguing for Copilot having no knowledge.

In other words, while you continue to push back against criticisms of your incorrect claims, without knowing it you are actually adding support to those very criticisms. I'm sure the irony of that escapes you, but I can't help but smile and shake my head. The sheer depth of the stupid this time around is mind boggling.
The universe does not work on "belief". It acts on information. It is the system (object) that has innate knowledge.
You lost track of what you were talking about, didn't you?

Copilot.... knowledge ... information ... what was I saying ... TEGMARK!

Can't you turn off that one track in your mind, even for a moment? Concentrate on what other people are telling you?

Nobody was discussing whether "the universe does not work on belief". Nobody but you. And you only started on that because you lost track of what the conversation is about, seemingly in the middle of posting a reply.

What's the matter with you, man? If you let us know what the problem is, you might even gain some sympathy. But this pretending there isn't a problem, while it's clear as day in every post you make, just sort of makes me feel sorry for you. Maybe you don't realise.
It always has correct information.
What? The universe? What were we talking about, again? Oh yes, Copilot.

Does Copilot always have correct information, Write4U? Are you seriously going to claim that it never makes a mistake or gives wrong information? Seriously? Do you want to try to defend that lost cause next?
There are plenty of organisms that have no self-awareness or beliefs, but they do have "innate knowledge".
We're not talking about organisms here. We're talking about an AI chatbot. Focus.
The formation of regular patterns suggests an ordering process?
You're not sure?
But a search engine does not discriminate between true and false either.
So you concede the point that Copilot is unreliable because it makes mistakes, then?
It is up to the investigator to separate the "chaff from the wheat".
So you concede that one should not outsource one's thinking to Copilot, now?
But that is not the correct question.
Focus on the point of contention.

It is your claim that Copilot is a search engine. DaveC claims it isn't. Sarkus claims it isn't. I claim it isn't. And - guess what! - it claims it isn't:

Sarkus: Are you a search engine?

Copilot:
Ah, I see you’re curious! While I’m not a search engine myself, I can certainly help you find information or answer any questions you might have.​

And your weak response to that is what? That asking it whether it is a search engine is the "wrong question". Instead, you claim, we should ask it "Can you be used as a search engine?"


With your query, it confirms that it is not a search engine. Or, maybe it is, just not "in the traditional sense", whatever that is supposed to mean.
It's not supposed to mean anything, because Copilot has zero, nil, nada, zip understanding of the text it is producing at any given moment. It merely has a mechanism for choosing the next word in its output. That's all. It has no beliefs. No knowledge. No store of facts. None of that. All it has is a large set of numbers that represent the relative weightings of what words most often come after what other words in a given context of English prose.

It follows that when Copilot writes "in the traditional sense", it doesn't mean anything to Copilot. For instance, it only generated the word "sense" because it's highly likely to come after the words "in the traditional" in English prose. And it only generated the phrase "in the traditional sense" due to contextual cues from the way the question was phrased and from its training data on how human beings have tended to answer that question on the web.
So, now we have two answers that seem to contradict one another: it says it is not a search engine, full stop. But it also says it is not a search engine in the traditional sense, which you want to read as it saying that it is a search engine in a non-traditional sense.

But Copilot can't be wrong, you tell us. So why would it contradict itself? Why would it give us contradictory answers to similar questions? It's almost as if it doesn't understand what it is being asked, isn't it? Funny, that.

Now, you have told us that Copilot has no beliefs. So it doesn't much matter that it says that it is a search engine and it says it isn't. It doesn't believe either of those claims.

Look at what else Copilot told you when it answered your query. It said it combines the functionality of a web search with AI tools. I note that a search engine does not combine the functionality of a web search with AI tools. Ergo, Copilot is not a search engine. This is consistent with its answer to Sarkus's direct question. (But we can safely ignore anything it says, anyway, using your criteria, because it has no beliefs.)

Who are you going to turn to, then, to learn about whether Copilot is a search engine or not? One option is to decide that Copilot itself is useless on that subject, because it contradicts itself. Another option is to decide that it's not entirely useless, because it arguably has some degree of consistency in its responses to similar question, but if you want to pay attention to it, then you'll notice that in all cases it comes down on the side of it not being a search engine, more or less.

Lucky for you, though. You have three actual human beings (presumably) right here on sciforums who do have beliefs, and guess what? None of us three believe that Copilot is a search engine. Then on the other side there's you, of course, digging that hole you're in deeper and deeper. What to do? What to do?

Keep on digging that hole of yours, I suppose. Right? You're comfortable in it. It seems to fit you. So roll with it.
It answered truthfully to the question as posited.
So did the humans. But you're going to trust the machine with no beliefs over the humans. Why? Because you made a silly claim and you can't bring yourself to admit you were wrong?
Ha, but that's not contradictory, it's discretionary.
You're not very good with this logic stuff, are you?
 
Last edited:
Here you are, agreeing that it is no beliefs. But knowledge is typically defined as justified, true, beliefs. So, you're arguing for Copilot having no knowledge.
I stipulated to that. OK, for now. It's just semantics.
It answered truthfully to the question as posited.
So did the humans. But you're going to trust the machine with no beliefs over the humans.
No I don't . That is your presumption.
Why? Because you made a silly claim and you can't bring yourself to admit you were wrong?
Where?
W4U, Ha, but that's not contradictory, it's discretionary.
You're not very good with this logic stuff, are you?
And that logical conclusion comes from what premise? You just throw ad hominem around as if you know what I am talking about. Let me assure you, you are way off.

Of the subjects under discussion, you know as little as I do. Nobody knows! That why these subjects are being hotly discussed by science.

And since I do not dispute established science, but use it like everybody else, that makes it even playing field on these subjects. I make no claim to any new "discovery" on my part.

I like a few of the newly introduced subjects and am trying to flesh them out in a general sense in order to gain some understanding. I don't need the minutia.

It's people like you who stand in the way of progress, locked up inside your own walls of old conventions.
 
Last edited:
Thanks James, for moving this thread. It is important enough to warrant its own thread.

I read once a statement by an AI developer: "Example, in a very complicated law case that requires deep research in prior law, the lawyer with an AI as research assistant will outperform the lawyer who is without, because of the extraordinary search capabilities of the AI."

So, I queried Copilot:in what way is copilot superior to conventional search engines?

Response:
You: in what way is copilot superior to conventional search engines?

Copilot
: Copilot takes all the functionality of a traditional web search and adds AI search engine tools to give you better, more thorough results in the formats you want. It can help you discover new things, learn new skills, and have fun along the way as you interact with conversational AI1.
Here are some ways in which Copilot stands out compared to conventional search engines: (highlights mine)

Natural Language Interaction: Copilot allows you to chat with it just like you would with a friend. Instead of typing in specific keywords, you can ask Copilot any question or explore any topic, and it will respond with helpful and engaging written answers. This conversational approach makes it easier to express complex queries and get relevant information.

Intent and Context Understanding
: Copilot is designed to understand your intent and context. You can ask multi-faceted questions, such as “Who is the president of France, and what is his net worth?” or “Show me some recipes for vegan lasagna and how to make them.” Copilot can handle these complex queries in one simple response, whereas traditional keyword searching might require multiple searches to get the same information.

Ongoing Conversation
: Copilot can follow up on your previous questions and provide additional information. With traditional search engines, you’d have to read through search results and start new searches for follow-up questions. Copilot offers near-instant, straightforward, AI-generated answers, and you can continue the conversation right there in the chat. Ask for more details, add specifics, or request further assistance.

Summarization
: Copilot excels at breaking down information. While traditional keyword searching leaves you to interpret and understand the content on your own, Copilot provides concise summaries and explanations.

Versatility Beyond Search
: Copilot’s functionalities extend beyond search. It can generate images, draft content, and work within documents directly in the browser. This versatility positions Copilot as a powerful tool that goes beyond what traditional search engines offer2.

In summary, Copilot combines the convenience of natural language interaction, context-awareness, ongoing conversation, and summarization to enhance your web browsing experience. Give it a try, and you’ll see how it simplifies finding answers to your questions!

I gave it a try, and its true! AI is not just a search engine, it is a "search engine plus".

If you only want information from peer reviewed papers, you can instruct AI to only search in peer reviewed papers.
 
Last edited:
It is and it's not.

I agree that it can skim through data fast and it can be targeted more specifically than you can easily do with Google. Google answers common quarries and if yours isn't common, you will likely just get the common response.

However, in cases where you don't get footnoted links with Copilot and in cases where you don't check them, you can get a lot of wrong answers. That's not generally the case with Google.

I use it (and others) a lot and it's very capable sometimes and sometimes it's very bad so you either need to already understand the subject matter or you need to use it as a first step and then confirm the accuracy yourself.

A lot of the errors are just from skimming and stopping at the first source and then misinterpreting that. If you correct it and then provide more specific information it will probably be accurate (or more accurate) this time.

Sometimes it's just a simple thing. I asked it if the Northern Lights would be visible from Seattle today. It said the chances were low according to NOAA but then it mentioned May 2nd.

I said, today is May 10th and it said the chances were high.

Other times it's an obvious math or logical error.

It's also annoying to the degree that it is programmed to give both sides to everything including a lecture, even when it's not a both sides issue.

Bard does that less but is generally much less capable. If you ask a long, detailed question or scenario it may just say "I'm still learning, I can't answer that". Yet, in some areas you can ask it for a prediction or something less than specific and it will try to give an answer with sources. Copilot will generally just say "consult with a professional" or something like that.

ChatGPT is pretty good for some things and less so for others.

You shouldn't assume that it's just a straight forward search engine that is rarely wrong however. It's wrong a lot more than a standard Google search. It is a good tool, a short cut to go through a lot of data and it's good for long, detail quarries that Google just doesn't do.

Just don't assume that the results are accurate without verifying. It's also good for a lot of other things. You could have a conversation with it and you might see a flaw in your logic (or not).

You could refresh your memory of books that you have read in the far past by just asking for a 1 or 2 page summary.

Most of the questions that you and Write 4 U ask, could be asked of Copilot initially such as questions about light, dark matter, dark energy, etc.
 
Last edited:
But that warranted the question if AI has access to peer reviewed papers.

So I searched.
Introducing Copilot: Your AI assistant that helps explain papers

Sucheth

May 25, 2023

"Here's one thing most of us agree on, processing information in research papers takes work. But wouldn't it be great if you could have AI explain papers to you no matter what the topic is, as you read, whenever you want?"
Picture this: every time you see a complicated equation in research papers, you’d get an explanation of it right there on the same screen. Even better, imagine you can ask any number of follow-up questions and get answers to them instantly. That's exactly what you get with SciSpace Copilot!
Use this AI assistant to get explanations and answers on any research paper as you read. Works for tables, equations, diagrams, jargon, and even lengthy blocks of text. You don't have to pause and search for it elsewhere. And your learning flow won't be disrupted.

Note: I am not trying to push Copilot per se. I'm sure other advanced AI are capable of interpreting similar complex search parameters.

AI that explains papers: Copilot for scientific research
For starters, the AI assistant is available across all the 200 million+ papers on the SciSpace repository. So, you can simply search for the paper or topic you want to look up to get started. Or, if you have the PDF stored on your device, you can sign up to SciSpace and then upload the same.
more....
https://typeset.io/resources/introducing-copilot-ai-assistant-explains-research-papers/
 
Last edited:
You shouldn't assume that it's just a straight forward search engine that is rarely wrong however. It's wrong a lot more than a standard Google search. It is a good tool, a short cut to go through a lot of data and it's good for long, detail quarries that Google just doesn't do.
Just don't assume that the results are accurate without verifying. It's also good for a lot of other things. You could have a conversation with it and you might see a flaw in your logic (or not).
I totally agree. Thanks for that cautionary tale.

AI is still in its infancy and as with all complicated systems, bugs are bound to crop up that need to be addressed.
My position is that in principle, the AI has greater research potential than dedicated search engines. No more, no less.

What is the best AI chatbot right now?
Thanks to its sourcing abilities, free internet access, and advanced LLM model, Microsoft Copilot is my choice for the best AI chatbot. Keep reading to see how its features compare to others like ChatGPT, You.com, and more.

The best AI chatbots of 2024
more.... https://www.zdnet.com/article/best-ai-chatbot/#The New Bing
and
What will you do with Copilot?
https://www.microsoft.com/en-us/bing?ep=466&form=MA13P3&es=31
 
OK, for now. It's just semantics.
Semantics is something you regularly struggle with. Maybe you should put a bit more time into working out the meanings of the words you use.
Of the subjects under discussion, you know as little as I do.
What subjects are you thinking of, specifically?
It's people like you who stand in the way of progress, locked up inside your own walls of old conventions.
What on earth are you talking about?

Which conventions am I supposedly locked inside of, Write4U?

If you're going to make silly accusations like that, at least try to back them up with some evidence or an argument for your position.

BTW, the age of a convention is not, in itself, a bad thing. We all use conventions all the time, some of them very very old indeed. That's not a flaw or a fault.

I suppose the point you're trying to make is that you think that you're somehow thinking outside the box, or breaking new ground in your revolutionary thought, whereas I'm so set in my ways that I just can't recognise the brilliance of your divergent thinking? Is that right?
 
If you're going to make silly accusations like that, at least try to back them up with some evidence or an argument for your position.
You don't back up your ad hominem with hard facts yourself. Practice what you preach.
 
BTW, the age of a convention is not, in itself, a bad thing. We all use conventions all the time, some of them very very old indeed. That's not a flaw or a fault.

BTW, the age of a convention is not, in itself, a bad thing. We all use conventions all the time, some of them very very old indeed. That's not a flaw or a fault.
Clearly you don't read my posts, else you would have noticed this.:
What subjects are you thinking of, specifically?
Microtubules for one. It is my dogged persistence that is responsible for the progress on this subject on this forum. I doubt you had even heard of them, when I introduced it as an important new area of scientific study.
And since I do not dispute established science, but use it like everybody else, that makes it even playing field on these subjects. I make no claim to any new "discovery" on my part.
 
Microtubules for one. It is my dogged persistence that is responsible for the progress on this subject on this forum.
That's the thing, though. There is no "progress" on that topic. You don't know any more about microtubules than you knew when you started posting about them, years ago. Not really. You developed an infatuation for the ideas of Hammeroff and Penrose, for some reason, and since then you've essentially just been engaged in a stamp collecting exercise, looking for the next pretty shell scientific paper that mentions the word 'microtubules', to add to your collection.

You don't understand much about most of the papers you randomly chop up and selectively quote on this forum. You're barely across even the definitions of some of the basic technical terms used in those papers. As far as I'm aware you have no formal training in biology or medicine, let alone microbiology. Without any of that, you can't really hope to make meaningful progress towards understanding any of the current research on microtubules. You have managed to delude yourself into imagining that using a search engine and the CTRL-C, CTRL-V keys on your computer amounts to "research" on the topic, but it really doesn't.
I doubt you had even heard of them, when I introduced it as an important new area of scientific study.
Microtubules were unknown to me until I read Penrose's book The Emperor's New Mind which I read soon after its publication in 1989. I'm guessing you arrived at that particular party later.

In case you are under some misapprehension, I have never claimed that microtubules are "unimportant". I am not a microbiologist, but I'm confident that there are microbiologists out there who find microtubules fascinating and who are expanding our collective knowledge of them. Certainly, they are a legitimate target for business-as-usual scientific research. As far as I am aware, research into them hasn't led to any earthshattering breakthroughs in biology or medicine, so far, but please correct me if you're aware of any.

I haven't seen any evidence that directly links microtubules to consciousness. Back in 1989, such a link was essentially blue sky speculation by Penrose and Hammeroff. The quantum aspects of microtubules that were touted as being vitally important for consciousness don't appear to have stood up to the scrutiny of the scientific establishment. In any case, no direct link between any putative quantum processes in the microtubules and consciousness has ever been established, to my knowledge. Certainly you did not post anything to support that notion, in years of posting on the topic, and this is supposed to be a special interest of yours.

Now, are you going to apologise to me for claiming that I am locked into restrictive conventions that make me unable to think creatively, or whatever it was you were trying to imply? Or have you already forgotten you posted that stuff?
 
Microtubules were unknown to me until I read Penrose's book The Emperor's New Mind which I read soon after its publication in 1989. I'm guessing you arrived at that particular party later.
Nope.
Penrose's book has no mention of microtubules. Penrose was "looking" for a vehicle that might be able to handle his non-algorithmic quantum wavefunction collapse that was necessary for his "consciousness" hypothesis.

It was Hameroff, in his capacity of anesthesiologist who was dealing with microtubules and consciousness, and had done serious research about their data processing capabilities.
Quantum Walks in Brain Microtubules—A Biomolecular Basis for Quantum Cognition?
tops12068-fig-0003-m.png


Stuart Hameroff
https://onlinelibrary.wiley.com/doi/full/10.1111/tops.12068

It was only after Hameroff read Penrose's book that he saw a possible solution to Penrose's dilemma.
Hameroff contacted Penrose and suggested that the microtubule might be what he was looking for. It was only then that Penrose became aware of the existence of microtubules and their possible capacity for data transmission at scales necessary for his concept. .

ORCH OR was the result of their collaboration, each contributing their expertise in his own field.
Actually Penrose was not so much impressed with neural microtubules as with cytoskeletal microtubules that allowed brainless cells to communicate with each other via "microtubules", r also known as "tunneling nanotubes" (TNT). Then there are also "microfibrils?

There seem to be several different names for microtubules, depending on the branch of cytology doing the research.
Perhaps one of the "learned fellows" can shed some light on this mystery.

Intercellular transport of Tau protein and β-amyloid mediated by tunneling nanotubes
In this study, we investigated the function of TNTs in the central nervous system. Using time-lapse imaging technology, we demonstrated that Tau protein and Aβ can be transported at different velocities in various cell lines through TNTs. The results provide a possible explanation for the cascade of nerve cell death in the brain during AD development. Furthermore, we found that lysosomes, mitochondria, and Golgi bodies can be transported by TNTs, suggesting that TNTs can mediate material transport between cells.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8661147/

Looks like a description of microtubules to me...:rolleyes:

Damn, this should be in the microtubule thread.
 
Last edited:
Nope.
Penrose's book has no mention of microtubules. Penrose was "looking" for a vehicle that might be able to handle his non-algorithmic quantum wavefunction collapse that was necessary for his "consciousness" hypothesis.

It was Hameroff, in his capacity of anesthesiologist who was dealing with microtubules and consciousness, and had done serious research about their data processing capabilities.

Actually Penrose was not so much impressed with neural microtubules as with cytoskeletal microtubules that allowed brainless cells to communicate with each other via "microtubules", r also known as "tunneling nanotubes" (TNT). Then there are also "microfibrils?

There seem to be several different names for microtubules, depending on the branch of cytology doing the research.
Perhaps one of the "learned fellows" can shed some light on this mystery.

Intercellular transport of Tau protein and β-amyloid mediated by tunneling nanotubes

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8661147/

Looks like a description of microtubules to me...:rolleyes:

Damn, this should be in the microtubule thread.
Wrong. "Tunnelling nanotubes" (TNTs) are not microtubules at all. They are tubes of cytoplasm connecting different cells, that allow the transport of cellular material between the two. Whereas microtubules are protein structures within a cell that contribute to the structural properties of the cytoplasm and are involved in cellular division etc. Microtubules are composed of a polymer of tubulin, a protein, and have a typical diameter of ~25nm. TNTs are membrane-bound cytoplasm structures - extensions of the cell in effect - with a diameter ranging from 0.05μ (i.e. 50nm) to 1.5μ (1500nm).

I was able to find this out by reading your link and then a Wiki article about TNTs: https://en.wikipedia.org/wiki/Tunneling_nanotube

Your post is a typical example, showing how little you understand about microtubules, in spite of deluding yourself that you have become kind of forum expert on the subject.

I don't know whether you used Copilot to come up with this irrelevance. If you did, it demonstrates that trying to use it as a search engine doesn't work when you have little understanding of the subject in question and limited ability to comprehend the content of what you are posting.
 
Your post is a typical example, showing how little you understand about microtubules, in spite of deluding yourself that you have become kind of forum expert on the subject.
Ok. I was not familiar with the name"tunneling nanotubes". Hence my question if anyone could explain.
Well, you explained with ad hominem expletives!

But I was not that far off. It looks like the two are functionally related. .

Microtubules regulate actin dynamics

https://www.bing.com/search?q=what is the relationship between actin and microtubules&pc=GD01&form=GDAVST&ptag=3601

And you are still acting boorish.
 
Last edited:
Ok. I was not familiar with the name"tunneling nanotubes". Hence my question if anyone could explain.
Well, you explained with ad hominem expletives!

But I was not that far off. It looks like the two are functionally related. .

Microtubules regulate actin dynamics


https://www.bing.com/search?q=what is the relationship between actin and microtubules&pc=GD01&form=GDAVST&ptag=3601

And you are still acting boorish.
There were no expletives.

Look, it's you who holds himself out to be some kind of forum expert on microtubules, not me. Yet in spite of all your hundreds of posts on the subject, you still know less about what they are than I do. That is pathetic.

Actin is a protein that forms microfilaments (NOT microtubules) within the cytoplasm, which enable the cell to perform certain actions. Microtubules also have a role to play in some of those processes. That does not make a microtubule a tunnelling nanotube.
 
Last edited:
And a little follow-up

Microtubules can transport actin filaments
We show that microtubules can transport actin filaments over large (micrometer-range) distances and find that this transport results from two antagonistic forces arising from the binding of cross-linkers to the overlap between the actin and microtubule filaments.


CYTOSKELETON
Cells contain a network of fibres made of protein. These fibres keep the cell’s shape stable by providing an internal framework called the cytoskeleton:
-->Some of the fibres, called microfilaments (made of actin filaments) are able to move against each other – these cause the movement seen in some white blood cells, and they move some organelles around inside cells. Movement is side-to-side like a wind-shield wiper.
-->There are other fibres, called microtubules. These are cylinders about 25nm in diameter made of a protein called tubulin, and may be used to move a microorganism through a liquid or to waft a liquid past a cell. Movement is circular like a helicopter propeller.

Comparing Microtubule & Microfilaments:
microtubule+and+microfilament2.jpg


Comparing Microtubule & Microfilaments:

http://thebiologs.blogspot.com/2013/11/cells-and-cell-structure-detailed-2013.html
 
Look, it's you who holds himself out to be some kind of forum expert on microtubules, not me. Yet in spite of all your hundreds of posts on the subject, you still know less about what they are than I do. That is pathetic.
Do you have some kind of inferiority complex that I should threaten your august status?
Where have I claimed expertise. I quote science because it is of interest to me.

I have specifically stipulated that I do not do science. I discuss science. That makes me a student, not a teacher.

Instead of encouraging me, for some reason you make me look like I am some kind of demonic existential threat to science itself.
Give it a rest, please.
 
Last edited:
Status
Not open for further replies.
Back
Top