|
|
Post by saratoga on Jan 24, 2023 21:49:55 GMT
Try chatting with ChatGPT:
I was surprised that it could converse in Japanese as well as in other languages. I expected this facility in its next evolution but it is already here.
When it starts to converse in spoken languages, there will be a further impact on language studies.
Reports say that it can now compose high school level papers. But I see a good possibility that it will become an essential tool for scholarly papers in maybe ten years time or less. I imagine that you can interact with it like you interact with a very competent and obliging coauthor. You throw out a topic and ask it to compose a draft. You discuss the draft with it and revise it and repeat. This way, you may finish a certain type of paper in a fraction of time it takes now. (By the way, this strategy is sometimes used by students using their professor as a human bot; the reverse is also true sometimes) Much of the effort would be in verification of results.
ChatGPT often gets facts wrong. I wonder where it gets these wrong information.
|
|
|
|
Post by marpro on Jan 24, 2023 22:37:55 GMT
Try chatting with ChatGPT:
I was surprised that it could converse in Japanese as well as in other languages. I expected this facility in its next evolution but it is already here.
When it starts to converse in spoken languages, there will be a further impact on language studies.
Reports say that it can now compose high school level papers. But I see a good possibility that it will become an essential tool for scholarly papers in maybe ten years time or less. I imagine that you can interact with it like you interact with a very competent and obliging coauthor. You throw out a topic and ask it to compose a draft. You discuss the draft with it and revise it and repeat. This way, you may finish a certain type of paper in a fraction of time it takes now. Much of the effort would be in verification of results.
|
|
|
|
Post by saratoga on Jan 26, 2023 22:12:15 GMT
saratoga: Do you think you can think as opposed to summarize others' opinions and reported facts?
ChatGPT: As a language model, I am not capable of thinking or having opinions as I am not a sentient being. I can only provide information and generate text based on the patterns and relationships in the data that I have been trained on. I can use my knowledge to generate new text based on the input provided, summarizing or providing a comprehensive answer based on the data that I have been trained on.
I can respond to prompts and provide information on a wide range of topics, but I am only able to provide you with the information that I have been trained on. I can't think or have opinions because I don't have consciousness or self-awareness.
I can assist you in providing information, generating text, answering questions or providing solutions based on the data that I have been trained on.
|
|
vl
Ensign
Posts: 18
|
Post by vl on Feb 15, 2023 3:36:28 GMT
What is the Ticker symbol for ChatGPT. Also what is the website
|
|
|
|
Post by racqueteer on Feb 15, 2023 12:11:11 GMT
What is the Ticker symbol for ChatGPT. Also what is the website It is an implementation of "AI"; an, imo, HUGE exaggeration as to its nature (not 'intelligent' so far as I can see). It's not an investment, so no symbol. Just do a search for it and start playing. It will rapidly sort through the accumulated 'knowledge' in existence (some of it incorrect and/or fabricated, unfortunately), and by whatever programmed algorithmic internal criteria, it will sort out what may be appropriate to your query. It's kind of like a quick search which sorts through the responses for relevance.
I'm not sure that you CAN invest directly; even on applications of the principles. It's also VERY early days; it's a curiosity at this point.
|
|
|
|
Post by steelpony10 on Feb 15, 2023 13:27:34 GMT
vl , I believe it comes out of a research facility. You can check this online. It is not a company but MSFT made a big investment in the research. If you invest in MSFT you just invested in that research.
|
|
|
|
Post by bobfl on Feb 15, 2023 13:33:42 GMT
Try chatting with ChatGPT:
I do want to play with it, perhaps test both the free and paid version. Could you try something? Ask it somehow to pick stocks. I realize it gets things wrong, but would be fun to track the performance of whatever it says in a pretend portfolio. |
|
|
|
Post by chang on Feb 15, 2023 13:43:45 GMT
Try chatting with ChatGPT: I do want to play with it, perhaps test both the free and paid version. Could you try something? Ask it somehow to pick stocks. I realize it gets things wrong, but would be fun to track the performance of whatever it says in a pretend portfolio. I just tried ChatGPT with a stock-picking question. The reply was:
As an AI language model, I cannot provide financial or investment advice. It is important to conduct thorough research and analysis before making any investment decisions.
When considering whether to invest in a particular stock, it's important to evaluate the company's financial health, growth prospects, competitive advantages, industry trends, and any potential risks or challenges. It can also be helpful to consult with a licensed financial advisor who can provide personalized advice based on your individual financial situation and investment goals.
Additionally, past performance is not a guarantee of future results, so it's important to consider a range of factors and not rely solely on historical stock prices or trends. |
|
|
|
Post by newtecher on Feb 15, 2023 13:59:37 GMT
I do want to play with it, perhaps test both the free and paid version. Could you try something? Ask it somehow to pick stocks. I realize it gets things wrong, but would be fun to track the performance of whatever it says in a pretend portfolio. I just tried ChatGPT with a stock-picking question. The reply was:
As an AI language model, I cannot provide financial or investment advice. It is important to conduct thorough research and analysis before making any investment decisions.
When considering whether to invest in a particular stock, it's important to evaluate the company's financial health, growth prospects, competitive advantages, industry trends, and any potential risks or challenges. It can also be helpful to consult with a licensed financial advisor who can provide personalized advice based on your individual financial situation and investment goals.
Additionally, past performance is not a guarantee of future results, so it's important to consider a range of factors and not rely solely on historical stock prices or trends.Must be hard coded to avoid investment advice to prevent lawsuits. I did have conversations with it about thermodynamics, pretending I had questions like a student would. I teach the subject and so know pretty well what students are likely to be confused about. About half of the time it gives vague but largely correct answers. But some of the time it gets specific and makes up outrageously wrong answers. It will be interesting to watch how it develops. Chances are that getting it to be reliable will be quite hard. Based on my experience, this tweet seems spot on. This recent article seems like a good way to think about large language models like ChatGPT. |
|
|
|
Post by bobfl on Feb 15, 2023 14:12:55 GMT
Thanks Admin〵Chang and New Teacher.
|
|
|
|
Post by saratoga on Feb 15, 2023 16:13:44 GMT
I find it very useful on health/medical questions. Of course, I do not trust the answers until I check them out with my doctors and other sources. But so far, all its answers seem believable.
I asked it to suggest very specific travel recommendations in Kyushu - an island in Japan. The answer looked like something created by an agent in US with little expertise on Japan but could serve as a starting point of a plan.
I asked it to give me great poems of ancient China (in Chinese characters) and Japan (in Japanese). It spat out some great poems. Once it balked and said that it could not write in Japanese. So, I said "But you could write it before." Then, ChatGPT apologized and answered in Japanese.
The AI is still incredibly useful so long as you are very careful about its limitations. Sorting and summarizing mountains of data in a few paragraphs is also an intelligence. There are symbolic math programs such as Maple, Mathematica as well as logic programs. Once AI integrates with them (not an easy task but doable, I think) there will be another big jump.
|
|
|
|
Post by newtecher on Feb 15, 2023 17:11:58 GMT
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 15, 2023 21:59:34 GMT
I just tried ChatGPT with a stock-picking question. The reply was:
As an AI language model, I cannot provide financial or investment advice. It is important to conduct thorough research and analysis before making any investment decisions.
When considering whether to invest in a particular stock, it's important to evaluate the company's financial health, growth prospects, competitive advantages, industry trends, and any potential risks or challenges. It can also be helpful to consult with a licensed financial advisor who can provide personalized advice based on your individual financial situation and investment goals.
Additionally, past performance is not a guarantee of future results, so it's important to consider a range of factors and not rely solely on historical stock prices or trends.Must be hard coded to avoid investment advice to prevent lawsuits. I did have conversations with it about thermodynamics, pretending I had questions like a student would. I teach the subject and so know pretty well what students are likely to be confused about. About half of the time it gives vague but largely correct answers. But some of the time it gets specific and makes up outrageously wrong answers. It will be interesting to watch how it develops. Chances are that getting it to be reliable will be quite hard. Based on my experience, this tweet seems spot on. This recent article seems like a good way to think about large language models like ChatGPT. So, exactly like most posters on forums! |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 15, 2023 22:15:24 GMT
I believe that it is important to note that AI (or machine learning or deep learning, for that matter) are all evolutionary and very dependent on speed (i.e. bandwidth) and hardware development. IOW, the ability to get to, and sort through, vast amounts of data is what defines the capabilities. And of course, the programming style. Having so much data migrating and growing on the web, with fast as light speeds (optical), and insanely fast processors is what is driving the whole thing.
This is just the beginning. The amount of new data available and the hardware advancements will just keep coming. Too soon to tell what it all will lead to, in terms of products and abilities. Where is Arthur C. Clarke when you really need him?
|
|
