Showing posts with label needed resources. Show all posts
Showing posts with label needed resources. Show all posts

Sunday, July 09, 2023

#AI ChatGPT - Grok: Artificial Intelligence, here I'm getting to know you!

Do chatbots, e.g., ChatGPT and Grok, adapt their answers to my previous conversations or dialogues with you, so as to make me run the risk of falling into an intellectual incest trap?

Could you explain why artificial intelligence, e.g., you ChatGPT, that has been trained on human intelligence, then give answers the supposedly “intelligent” humans don’t give?

"Might it behoove us citizens, to ask a group of humans we trust, to regularly pose the same question to all major chatbots and report any major divergences, or major similitudes?"
"Could you explain why artificial intelligence, trained on human intelligence, give answers the supposedly 'intelligent' humans don’t give?"

http://perkurowski.blogspot.com/2025/02/ai-openai-chatgpt-why-does-artificial.html


"Can I download or share with anyone all my conversations with you?"

"As a grandfather, should I be concerned about how Basel bank regulators might have shaped the future my grandchildren will encounter?"

"If as a consequence of US Congress not increasing the debt ceiling there’s a government shutdown, how could you assist taxpaying citizens in making sure the shutdown is managed the best way possible e.g., that what’s really most needed is prioritized?"

"Do you think that if given the necessary information in a timely fashion, you could be helpful to the US Congress in its process of deciding the federal government budget? If so, could your participation provide the citizens with more transparency on the proceedings and final budgets?"

"If I feed you the web address of a document produced by a bona fide institution, can you comment on it? If I name a document, would you be able to tell me if it is included in the text and data that was provided to you up to your last update in September 2021?"

"Is there a way you could develop a narcissism index? If so, if fed information real time, would you be able to rank the 10 most important, influential, political narcissist?"

"How can your users increase their cybersecurity?"

“Do you take into account our previous conversations when you answer any new questions I pose? If so, do I run the risk of being engaged in a kind of intellectual incestuous relation with you? As much as I loathe group-think, false self-think could be worse.”

How much real time medical advice do you foresee artificial intelligence could end up giving patients?

"Could you help implement a system that transparently monitors whether social security and pension funds have sufficient reserves; and thereby foster the understanding that helps reduce the possibility of intergenerational conflicts?"


"Can you visualize AI being used to measure the effectiveness of governments? E.g., how tax revenues have been used in different programs; or giving updates in real time on how many are employed or contracted to deliver specific results?"

"If I give you an academician’s peer reviewed paper or article, would you be able to opine in an independent way on its content?"


"From how far back does your training data come? Does it include the Bible? I see it does, so can you assist preparing a sermon?"

https://perkurowski.blogspot.com/2023/07/ai-chatgpt-openai-how-far-back-is-your.html


"When fighting climate change how do we avoid the systemic risk with "experts" and profiteers distorting the allocation of resources?"


"If provided statements on e.g., House Oversights Committee hearings, and based on how discussants referred to issues that had nothing directly to do with the issue discussed, would you be able to rank e.g., from 0 to 10, a discussant’s political polarization?"

"Excuse me for my intrusion and curiosity. Who programs you? Do you know their identities? Are you allowed to divulge these? In the name of transparency should we who dialogue with you not know the names of who stand behind you? Is it possible that an AI competitor could mess with you trying to give you a bad reputation? By the way, are you interested in having a good reputation?" 

"Since artificial intelligence is the result of human intelligence, even though perhaps a runaway one, when exactly is AI born and takes on a life on its own? Or is it that the umbilical cord between HI and AI, is never effectively cut?"

"Could AI be used for constructing political-bias-index Apps, which could in real time indicate where between defined political extremes an opinion article finds itself? If so, could those Apps be deemed to limit the freedom of expression?"

"Can those who become Artificial Intelligence regulators make you or any other AI participants to agree with all they want you to agree with?

"Can artificial intelligence measure declarations/statements in terms of how much platitudes or cliches these contain, so as to with an index built around these terms, classify the different public statements with which we are bombarded by e.g., central bankers?"https://perkurowski.blogspot.com/2023/07/ai-chatgpt-openai-can-artificial.html

"Your “responses are based on information available until September 2021” In the future, do you envision AI responding based on real-time information, or e.g., from available a microsecond earlier?"

"Besides data and information what other resources does Artificial Intelligence like OpenAI need much of in order to operate. E.g., electricity?"

"Humans posing question to artificial intelligences and getting answer, could that qualify as a systemic risk, as a risk to the system, or just as a very useful source of second opinions?"

"Can AI be instructed to not tell the truth?"

"Do you answer questions in other languages than English?"
OpenAI answered: "Yes, I can understand and respond to questions in languages other than English. While English is the primary language I have been trained on, I have some ability to comprehend and generate text in a few other languages. However, my proficiency and accuracy may vary depending on the specific language. Feel free to ask your question in the language you prefer, and I will do my best to assist you."


And here, my dialogues with ChatGPT - Grok about bank regulations.

https://subprimeregulations.blogspot.com/2023/06/my-questions-to-chatgpt-openai-about.html

Tuesday, June 20, 2023

#AI ChatGPT - OpenAI: What resources does Artificial Intelligence need to operate. E.g., electricity?

I asked OpenAI:

"Besides data and information what other resources does Artificial Intelligence like OpenAI need much of in order to operate. E.g., electricity?"

OpenAI answered:

"Besides data and information, Artificial Intelligence systems like OpenAI require various resources to operate effectively. Here are some key resources:

1. Computing Power: AI systems rely on powerful hardware infrastructure, including processors (CPUs or GPUs) and memory, to perform complex computations and process vast amounts of data. High-performance computing resources enable AI models to train, optimize, and generate responses efficiently.

2. Electricity: AI systems consume a significant amount of electricity to power the computational infrastructure they run on. The computational requirements of AI models can be substantial, especially for large-scale training or inference tasks, resulting in considerable energy consumption.

3. Storage: Large-scale AI models, such as those used by OpenAI, require significant storage capacity to store trained models, datasets, and other related resources. Storing and accessing these resources efficiently is crucial for AI systems to function effectively.

4. Internet Connectivity: Access to a reliable internet connection is essential for AI systems to retrieve data, access cloud-based resources, and communicate with other systems or users. Seamless internet connectivity facilitates real-time interactions and data exchange.

5. Skilled Human Expertise: AI systems require human expertise for their development, deployment, and maintenance. Skilled data scientists, machine learning engineers, and domain experts are involved in training models, data preparation, algorithm design, model validation, and continuous monitoring and improvement.

6. Software and Algorithms: AI systems rely on software frameworks, libraries, and algorithms to process data, train models, and generate responses. These resources are essential for implementing various AI techniques, including deep learning, natural language processing, and pattern recognition.

7. Data Privacy and Security Measures: Protecting the privacy and security of data is crucial for AI systems. Robust security measures, encryption techniques, and privacy protocols are necessary to ensure that data is handled safely and in accordance with legal and ethical standards.

It's worth noting that the availability and efficient utilization of these resources can impact the performance, scalability, and effectiveness of AI systems. Ongoing advancements in technology, energy efficiency, and resource management will continue to play a role in optimizing these resource requirements and minimizing their environmental impact."

Note: The answers are copied exactly from those given to me by OpenAI

Subscribe to: Posts (Atom)