To what extent does the ChatGPT technology independently learn to improve the answers given to the questions asked?
To what extent does the ChatGPT consistently and successively improve its answers, i.e. the texts generated in response to the questions asked, over time and when receiving further questions using machine learning and/or deep learning?
If the ChatGPT, with the passage of time and the receipt of successive questions using machine learning and/or deep learning technology, were to continuously and successively improve its answers, i.e. the texts generated as an answer to the questions asked, including the same questions asked, then the answers obtained should, with time, become more and more perfect in terms of content and the scale of errors, non-existent "facts", new but not factually correct "information" created by the ChatGPT in the automatically generated texts should gradually decrease. But has the current, next generation of ChatGPT 4.0 already applied sufficiently advanced, automatic learning to this tool to create ever more perfect texts in which the number of errors should decrease? This is a key question that will largely determine the possibilities for practical applications of this artificial intelligence technology in various fields, human professions, industries and economic sectors. On the other hand, the possibilities of the aforementioned learning process to create better and better answers to the questions asked will become increasingly limited over time if the knowledge base of 2021 used by ChatGPT is not updated and enriched with new data, information, publications, etc. over an extended period of time. In the future, it is likely that such processes of updating and expanding the source database will be carried out. The issue of carrying out such updates and extensions to the source knowledge base will be determined by the technological advances taking place and the increasing pressure on the business use of such technologies.
In view of the above, I address the following question to the esteemed community of scientists and researchers:
To what extent does ChatGPT, with the passage of time and the receipt of further questions using machine learning and/or deep learning technology, continuously, successively improve its answers, i.e. the texts generated as a response to the questions asked?
To what extent does the ChatGPT technology itself learn to improve the answers given to the questions asked?
What do you think about this topic?
What is your opinion on this subject?
Please respond,
I invite you all to discuss,
Thank you very much,
Warm regards,
Dariusz Prokopowicz
GPT-4, the latest incarnation of the artificial-intelligence (AI) system that powers ChatGPT, has stunned people with its ability to generate human-like text and images from almost any prompt. Researchers say this type of AI might change science similarly to how the Internet has changed it. Yet many people are frustrated that the model’s underlying engineering is cloaked in secrecy. Assurances about GPT-4’s improved safety by its creator OpenAI fall short for some. “It’s just completely impossible to do science with a model like this,” says AI researcher Sasha Luccioni...
https://www.nature.com/articles/d41586-023-00816-5
GPT-4 is even more transformative for creating media. It can handle 25,000 words! It can easily work with you to write books, compose songs or research papers. With 25,000 words AI will be much more powerful. For example, you can ask it to consume all of David Ogilvy’s books on copywriting and use what it’s learned to rewrite landing pages, sales pages, your home page. No person can keep 25,000 words in their head, and effectively reason and manipulate with the information. The impact here will be incredible. It will transform education. I can give GPT-4 whole sets of documents and ask it to provide a step-by-step guide to teach me what’s in it …… woah. Khan Academy just launched an AI tutor for their courses...
https://www.linkedin.com/posts/kieranjflanagan_last-night-we-all-got-new-superpowers-gpt-activity-7041763466428194817-ZJY2/?utm_source=share&utm_medium=member_desktop
It is logical consequence!
Governments worldwide attempting to regulate generative AI
Countries worldwide are monitoring the generative systems with their own unique rules and laws. For example, China proposed new laws, and the U.S. requested public comments...
The potential impact of regulation on how the technology is both developed and deployed will take some time to resolve...
https://www.techtarget.com/searchenterpriseai/news/365535037/Governments-worldwide-attempting-to-regulate-generative-AI?_gl=1*1vdlijo*_ga*MTU2OTQ0MjI1My4xNjc3NzUwODI2*_ga_TQKE4GS5P9*MTY4MTcxODk4MC4yNC4xLjE2ODE3MTkxNzkuMC4wLjA.*_ga_RZDF13FDNT*MTY4MTcxODk4MC4yMi4wLjE2ODE3MTg5ODAuMC4wLjA.*_ga_NLDTRJGG3Y*MTY4MTcxODk4MC4yMC4wLjE2ODE3MTg5ODAuMC4wLjA.*_ga_H4TNQB84WS*MTY4MTcxODk4MC4yMC4wLjE2ODE3MTg5ODAuMC4wLjA.*_ga_7FK328ZGNW*MTY4MTcxODk4MC4xOS4wLjE2ODE3MTg5ODAuMC4wLjA.&_ga=2.134786801.1201627196.1681718980-1569442253.1677750826
Dear Dariusz Prokopowicz , this is the example of chatGPT made interview.
What artificial intelligence asked Miroslav Mišković: ChatGPT was asked to conduct an interview with the owner of Delta...
One may use Google Translate to read it.
https://www.danas.rs/vesti/ekonomija/sta-je-vestacka-inteligencija-pitala-miroslava-miskovica-od-chatgpt-a-zatrazeno-da-uradi-intervju-sa-vlasnikom-delte/?utm_source=mailpoet&utm_medium=email&utm_campaign=pregled-29-04-2023
Microsoft has opened up the OpenAI-enabled features in Bing, which were previously available as a limited preview. The expanded preview comes just days after the company rolled out new pricing for Bing API, which has seen a licence fee hike for developers who want to use Bing’s service in their own systems...
The company aims to cement artificial intelligence (AI) in Bing, moving from text-only search to offering results that also embed images and video answers...
https://www.computerweekly.com/news/366536522/Bing-AI-push-rings-in-price-change?utm_campaign=20230505_Bing+AI+push+rings+in+price+change&utm_medium=email&utm_source=MDN&source_ad_id=-1&asrc=EM_MDN_266755553&bt_ee=26KG%2FWvO19GG5WutYwVhxysZA7KpYoVDWeiGwbuo7K9s9qguEwvgbqYGkHhhwWC%2BxawwV4oHuYZ7RNUWoOg9iA%3D%3D&bt_ts=1683455355918
AI-Based Chat Powers an Improved Customer Experience
The AI-based chat system provides quick resolution to issues, either automatically or via experts...
Those of us who have been trapped in an endless loop with a poorly programmed chat system can be comforted by knowing advancements in AI are rapidly evolving to improve the digital experience at scale. The benefit and power of such tools and communication strategies is evolving daily, and users can now select self-service options regarding measurement technologies and applications for immediate answers...
https://www.automation.com/en-us/articles/may-2023/ai-based-chat-powers-improved-customer-experience
Yes, it definitely happens, indeed, within a week of talking to one of these, you can turn this artificial intelligence creature into a minimally responsive professor in any of the specialized fields! Not because we know a lot!, but with the questions we ask him, we subject him to many trainings, and he learns these trainings by wandering through huge databases according to complex algorithms! I don't know if this is good for humanity in an operational and field situation!? Definitely, only those who will benefit from the performance of these new artificial intelligence creatures, who will set a precise balance between their individual efforts in trying to solve problems and handing over the solution of these problems to them. If we think carefully, the feeling of danger caused by this situation is really serious!
https://www.newyorker.com/science/annals-of-artificial-intelligence/there-is-no-ai
''The most pragmatic position is to think of A.I. as a tool, not a creature.''
"Lanier is diplomatic here, he has too many friends in Silicon valley so he do not want to offend then and undiplomatically laugh at the ludicrous idea of 'A.I. as a creature"...
To consider A.I. as a creature is "pure Machiavelian pragmatism"...
While generative AI such as ChatGPT gets all the buzz, AI in other, typically more targeted forms, has been used for years to solve business and research challenges. The utility of AI and the value to businesses is directly related to both the quality of inputs and information about what inputs are used. Even when an AI service is trained on high quality content, without proper documentation and audit trail, users cannot be sure what they are using and may stay away...
And while in-context learning currently shows promise for reducing the amount of data needed for retraining, high-stakes uses will still need higher value, transparently documented content for training the large language models themselves.
While large data sets scraping the web (often without consent) are all the rage, the use of high quality, documented data will be important in advancing science...
https://scholarlykitchen.sspnet.org/2023/05/15/swimming-in-the-ai-data-lake-why-disclosure-and-versions-of-record-are-more-important-than-ever/?informz=1&nbd=6f03e560-5431-4744-8998-e00223ee7a82&nbd_source=informz
This is an in-depth comparison of ChatGPT vs Google Bard
ChatGPT and Google Bard are AI chatbots designed to generate responses to prompts. When used appropriately, ChatGPT and Google Bard can be used to support certain business processes in content production, development and more. Take a look at each tool’s features, pros and cons to see which would be best for your business...
https://www.techrepublic.com/article/chatgpt-vs-google-bard/
Scientists disagree on whether artificial intelligence can reason and are struggling with how to assess its capabilities. GPT-4 and other large language models ace academic exams and would probably breeze through a conversation-based Turing test. At the same time, they have glaring blind spots, such as failing an exam question when it is slightly changed, and they seem to lack abstract reasoning abilities. “There’s no Geiger counter we can point at something and say ‘beep beep beep — yes, intelligent’,” says cognitive scientist Tomer Ullman...
https://www.nature.com/articles/d41586-023-02361-7
Large language models (LLMs), a type of artificial intelligence (AI), can write poetry, code and pass tough exams. Often, the precise reasons why they behave the way they do are not known — even to their own creators. “Revealing their true nature is urgent and important,” argues a Nature editorial. To really assess LLMs’ abilities and limitations, researchers need to understand how they work and find new ways of testing them. “Similar to the processes that medicines go through to attain approval as treatments and to uncover possible side effects, assessments of AI systems could allow them to be deemed safe for certain applications,” the editorial suggests...
https://www.nature.com/articles/d41586-023-02366-2
Until it is connected to the internet, Chat GPT is a really big brain stuck in a room.
So instead of reading “we fed ChatGPT” think we fed “big brain in a room” the following prompt: I am a research scientist in NLP area. I am currently interested in large language models like GPT-3. Can you recommend some relative papers about this area?
https://scholarlykitchen.sspnet.org/2023/08/30/guest-post-was-chatgpt-set-up-to-fail-choosing-the-right-tools-and-the-right-prompts-is-essential-for-llm-discovery/?informz=1&nbd=6f03e560-5431-4744-8998-e00223ee7a82&nbd_source=informz
Going forward the number of word responses as well as quality of the answers will improve given that ChatGPT is AI powered.
Dear Dariusz Prokopowicz , this is very good article:
What Is ChatGPT Doing … and Why Does It Work?
That ChatGPT can automatically generate something that reads even superficially like human-written text is remarkable, and unexpected. But how does it do it? And why does it work? My purpose here is to give a rough outline of what’s going on inside ChatGPT—and then to explore why it is that it can do so well in producing what we might consider to be meaningful text...
https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-doing-and-why-does-it-work/
What is ChatGPT not?
Talking to ChatGPT can feel remarkably like talking to something sentient. And in some sense, the chatbot is intelligent: it can do some tasks, such as passing exams, as capably as a human. Yet it fails simple logic puzzles that people find easy. It’s important to remember that these systems aren’t independent agents: they have no intentions and no concept of what is true or false...
https://www.nature.com/articles/d41586-023-04156-2
- Badges
- Science topic