Haven’t done one of these in a while. We recently passed 8,000 subscribers on TAB! so I decided it’s time for another open AMA session to celebrate!!As usual, you can ask me about pretty much anything: AI (current trends, predictions, companies, models, etc.), writing, Substack, my background, future plans… whatever!
And, of course, about The Topic—ChatGPT. What are my thoughts, its virtues and flaws, the consequences and implications across areas, in education or industry, and its role in the future of AI.
Bring in the questions. I’ll try to answer all of them!
Is ChatGPT and similar a psychological, emotional help to lonely people? And if so, will more interaction between the lonely people and the ChatGPT lead to despair?
Will European countries drop out of the AI technology improvement processes while they search for a way to apply some sort of regulation on their local industry looking to work with AI technologies? What do you expect will be their ability to offer competitive products and services while they are studying how they want to regulate it in Europe?
If you could create an robot/AI for any task, small or large (answering your phone and creating calendar events/clothes-shopping accurately for you) what task would you automate?
Hi My Name is Claire, I own creative small business. My mission this year is to help creative to stay informed about the advancement of generative Ai including GPT 3 ChatGPT, as well as generative Images, audio, video tools etc. My concern right now mostly people are actually learning and adopting are from the tech world of the bubble. I’ve feeling the a little pushback from my creative community. and feeling kind alone in my mission that I can’t find an other person who’s from creative community is passionate about keep creatives up to date so we can stay competitive for the future of the creative industry. How should I approach this, and what kinda content should I be sharing and creating that will help me in this journey.
I already started post on my Instagram and TikTok pages, and I put together a list of resource, people tools to follow on motion called The Ai Stuff Creator and Artist Should know.
What are the most effect way that I help to spread out the word?
Many technologies, possibly most, define themselves around one core application. Generative AI has certainly avoided this temptation. Perhaps it will continue to do so. If it does not I have a suggestion. .I think in ten years we will see that the dominant application of the technology will be in allowing humans to have robust, fully interactive, conversations with machines. Machines of all sorts and types and functions. To take an example, perhaps an extreme one, imagine being able to talk with your house about your needs and imagine that it had useful reactions. It will be like having one of those old family servants known to us from the writings of Dickens.
I am currently writing a series of articles about AI for a Danish healthcare research media. One of them digs into the hype vs. reality question since there have been a lot of predictions about the stunning things AI can do for the health care sector. However, in practice the field has a hard time living up to these promises. Research studies showing great results rarely works all that well in reality. Is this a general problem for AI or is it especially outspoken within healthcare research?
Congrats on the 8,000 subs Alberto! Way to go. Jealous!!
Here's a menu of questions to give you some choices. Answer any that interest you.
QUESTION 1: What are the compelling benefits of AI which justify introducing what could be another existential risk? Or, if you prefer, what are the compelling benefits of AI which justify providing psychopaths like Putin with even more power?
QUESTION 2: Is the marriage between violent men and an accelerating knowledge explosion sustainable?
QUESTION 3: Why do I have literally 1,000 times fewer subscribers than you? :-)
I am personally quite unimpressed with ChatGPT, when its results are compared to what a human being can produce. However, it is understandable that educational institutions are worried about its possible misuses, particularly in the context of high-school assignments, which ChatGPT can easily reproduce (since all they require is the summarizing and regurgitating of widely unambiguously accepted knowledge).
I have a few questions regarding these issues, they are all quite general:
1) Could you comment on strategies which could employ tools such as ChatGPT for enhancing education, rather than seeing it as a danger to education?
2) Could you comment on issues of style (e.g. the fact that in e.g. academia a particular kind of "disengaged" style is promoted, while in fact what will ultimately distinguish ChatGPT products from human products is precisely a question of 'unique' style)?
3) It is likely that, because of the willing adoption of tools such as ChatGPT, the production of text will drastically increase (making a lot of journalistic/editorial/etc. work obsolete), but because of this it is also likely that the need for curatorial and summarizing services will increase as well. What this means, abstractly, is that there will be massive amounts of texts being produced from "basic units" of info (e.g. "write a friendly email to my coworkers that I will not make the deadline"), which at the same time, on the other end, become summarized back into these "basic units" because people don't have time to parse through vast seas of "useless" grammatization (e.g.: "write me a summary of all the incoming email regarding the project deadline"). This leaves me wondering, not whether there is any objective reality to the "basic units" (which I don't believe there is), but where the belief in an objectively summarizable (and, at the same time, expandable yet expendable) reality comes from. Perhaps the question is even this one: where does the biased focus on language as a tool come from? (Perhaps it is a vestige of capitalist product-oriented obsessions?) Why, when we are impressed by the poetic abilities of ChatGPT, can we not also appreciate language in general as the basic social fabric that holds our relational dynamics together? The "bias" discussion on the partiality of systems seems not to have taken actual critical hold, because we continue to promote systems which supposedly parse reality in "neutral" and unbiased ways. This last comment/question is quite abstract, but I hope you're able to comment on it.
1) Have you heard of Rain neuromorphics?
Recently, I have become interested in neuromorphic chips. If systems like ChatGPT have caused so much horror and admiration, then what will happen when neural networks on neuromorphic chips appear? Here to find out what you think about this topic?
2) What do you think about the "forward forward" algorithm?
3) What do you think about the possible appearance of a "useless class"? People will always be needed, but in the near future probably only certain ones, but which ones do you think?
When will ChatGPT be able to do my laundry?
What is this biggest misconception about AI?
What are some mind-blowing capabilities you think a multi-modal gpt-4 will open up? How quickly/competently do you see Google responding to OpenAI/Microsoft with their own products?
Thanks for your work. It's really helpful because it's well informed, well sourced and nuanced. I also follow G. Marcus on Substack.
I'm a French law librarian researcher and a lawyer by education and I monitor developments of the use of AI (machine learning, LLM, expert systems) for law.
I wonder if you have feedbacks and info from the legal community on ChatGPT, GPT and the like.
My two questions are :
- do you think generative IAI's hallucinations can be reduced to a point where they won't matter to professionals? In law, that's a key question
- are there any signs that LLM will be able to reason? Legal reasoning is done with words, I know, so in theory do you think it could be done?
My (long) post (in French) on the subject is here (and on the ChatGPT matter it owes you some) : https://www.precisement.org/blog/Intelligence-artificielle-en-droit-derriere-la-hype-la-realite.html
Hi. Sorry for being late to this ama. But I have a few questions. 1. With regards to having a foundational understanding of AI, how crucial it is to have good knowledge of statistic. 2. With regards, to many generative AI applications. Do you think these guys train the model from scratch or they start with some pre-trained model.3. Lastly, with regards to combining ChatGPT or LLM in general with search algorithms, what do you think is the approach taken by Google and Microsoft. Do they create a new models that incorporate both LLMs and search algorithms, and in general , does mixing this two affect the accuracy and relevancy of the search results
Greg as I understand nothing stops you for using online sources but there are likely guidelines on how you give credit to the sources you used. There is also a really important role that you’re always responsible for what is said and the quality/accuracy of what you write under your name as author. So keep that in mind with any sources you use to make sure you are as accurate as possible! Don’t ever count on a single source to be always correct.
A.I. Can help find truth. That saves energy. Test internal reality against A.I. statements of fact. Great new future for machine truth.
What are the practical implications of the difference between using ChatGPT as tool vs playing with it as a partner?
What sort of AI model is best suited to track and predict human behaviour?
Hi, what are the environmental impacts of AI?
How do or do ChatGPT and its like affect machine translation? Any hope for them as editors?
How will universities combat the use of chatgpt?