[CG] Re: Insights into three of the important problems associated with LLMs/Generative AI

Thanks for your wise words on the limitations of LLMs, John. Re: any mission-critical application. It is best used under the control and supervision of some system that uses trusted methods of AI and computer > science to check, evaluate, and supplement whatever the generative AI happens to generate. Another source of "generative AI control and supervision" may actually be human collectives. Case in point: I am in the "Deliberation At Scale" consortium, one of the winners of the OpenAI "Democratic Inputs to AI" grants: https://openai.com/blog/democratic-inputs-to-ai In our consortium, we use ChatGPT as a moderator in scalable discussions on wicked problems like climate change (or in the pilot case: "how to make AIs more ethical?"). Chatbots help facilitate and summarize our human group discussions. Each group then decides whether a bot-proposed summary captures the essence of their discussion before it is being passed on to the next group for further commenting etc. It's a very interesting emerging symbiosis of human and machine intelligence, much more alike to Doug Engelbarts on AUGMENTATION than mere automation of human intelligence. So perhaps we'll see ever more of these human-machine amalgams to rein in runaway AI? Aldo On Wed, Oct 11, 2023 at 9:53 PM John F Sowa &lt;sowa(a)bestweb.net&gt; wrote: > Amit and anybody who did or did not attended today's talk at the Ontology > Summit session, > > All three of those questions below involve metalevel issues about LLMs and > various reasoning issues with and about generative AI. The first and most > important is about anything generated by LLMs: Is it true, false, or > possible? After that are How? Why? and How likely? > > The biggest limitation of LLMs is that they cannot do any reasoning by > themselves. But they can often find some reasoning by some human in some > document from somewhere. If they find something similar, they can apply it > to solve the current problem. But the word 'similar' raises critical > questions: How similar? In what way is it similar/ Is that kind f > similarity relevant to the current question or problem? > > For example, the LLMs trained on the WWW must have found textbooks on > Euclidean geometry. If some problem is stated in the same terminology as > the books on geometry, the LLMs might find an answer and apply it. > > But more likely, the problem will be stated in terms of the subject > matter, such as building a house, plowing a field, flying an airplane, or > surveying the land rights in a contract dispute. In those cases, the same > geometrical problem may have few or no words in common with Euclid's > description of the geometry and the terminology of each of the applications. > > any mission-critical application. It is best used under the control and > supervision of some system that uses trusted methods of AI and computer > science to check, evaluate, and supplement whatever the generative AI > happens to generate. > > As an example of the kinds of systems that my colleagues and I have been > developing, see https://jfsowa.com/talks/cogmem.pdf , Cognitive Memory > For Language, Learning, and Reasoning, by Arun K. Majumdar and John F. > Sowa. > > See especially slides 44 to 64. They show three applications for which > precision is essential. There are no LLM systems today that can do > anything useful with those applications or anything similar. Today, we > have a new company, Permion.ai LLC, which has developed new technology that > takes advantage of *BOTH *LLMs and the 60+ years of earlier AI research. > > The often flaky and hallucinogenic LLMs are under the control of > technology that is guaranteed to produce precisely controlled reasoning and > evaluations. Metalevel reasoning is its forte. It evaluates and filters > out whatever may be flaky, hallucinogenic, or inconsistent with the given > facts. > > John > > ------------------------------ > *From*: "Sheth, Amit" &lt;AMIT(a)sc.edu&gt; > > There has been a lot of discussion on LLMs and GenAI on this forum. > I would like to share papers related to three major challenges: > > 1 Is it Human or AI? d > Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as > You May Think — > Introducing AI Detectability Index > > 2. Measuring, characterizing and countering Hallucination (Hallucination > Vulnerability Index) > The Troubling Emergence of Hallucination in Large Language Models –An > Extensive Definition, Quantification, and Prescriptive Remediations > > 3. Fake News/misinformation > FACTIFY3M: A Benchmark for Multimodal Fact Verification with > Explainability through 5W Question-Answering > > Introduction/details/links to papers (EMNLP 2023): > https://www.linkedin.com/feed/update/urn:li:activity:7117565699258011648 > > I think this community won’t find this perspective alien: > > Data driven only approaches can’t/won’t address these challenges well— > need to understand the duality of data and knowledge > <https://arxiv.org/abs/2103.13520>. > Knowledge (including KGs/ontologies/world model/structured semantics) and > neuro-symbolic AI <https://arxiv.org/abs/2103.13520> (arxiv > <https://arxiv.org/pdf/2305.00813.pdf>) which use a variety of relevant > knowledge (linguistic, common sense, > domain specific, etc) will play critical role in > addressing these. The same goes for three of the most important > requirements > (knowledge will play a critical role in making progress on these): > grounding, intractability, and alignment. > More to come on this from #AIISC. > > Cheers, > Amit > > Amit Sheth LinkedIn, <https://www.linkedin.com/in/amitsheth/>Google > Scholar, > <https://scholar.google.com/citations?user=2T3H4ekAAAAJ&hl=en&oi=ao> > Quora, <https://www.quora.com/profile/Amit-Sheth-1>Blog, > <https://amitsheth.blogspot.com/>Twitter > <https://twitter.com/amit_p?lang=en> > Artificial Intelligence Institute; NCR Chair > University of South Carolina > #AIISConWeb, <https://aiisc.ai/>#AIISConLinkedIn, > <https://www.linkedin.com/company/aiisc/mycompany/>#AIISConFB > <https://www.facebook.com/AIIUofSC> > > > _______________________________________________ > CG mailing list -- cg(a)lists.iccs-conference.org > To unsubscribe send an email to cg-leave(a)lists.iccs-conference.org > -- ========================================================================== Aldo de Moor, PhD CommunitySense - for working communities Cypresstraat 24, 5038 KS Tilburg, the Netherlands e-mail: ademoor(a)communitysense.nl mob: +31-6-47011400, tel/fax: +31-13-4564126 site: www.communitysense.nl KvK: 18088302 blog: communitysense.wordpress.com mastodon: https://mastodon.social/@ademoor twitter: ademoor ==========================================================================