{"id":2994,"date":"2023-01-29T14:26:14","date_gmt":"2023-01-29T08:56:14","guid":{"rendered":"https:\/\/www.rangakrish.com\/?p=2994"},"modified":"2023-01-30T18:51:45","modified_gmt":"2023-01-30T13:21:45","slug":"wolframalpha-chatgpt-and-the-future-of-ai","status":"publish","type":"post","link":"https:\/\/www.rangakrish.com\/index.php\/2023\/01\/29\/wolframalpha-chatgpt-and-the-future-of-ai\/","title":{"rendered":"WolframAlpha, ChatGPT, and the Future of AI"},"content":{"rendered":"

We all know that ChatGPT<\/strong><\/em><\/a> has taken the world by storm. True, it is a major advance of Artificial Intelligence in the area of Natural Language Processing. Many may not know that WolframAlpha<\/strong><\/em><\/a>, launched in 2009, allowed natural language queries. As a long time user of Wolfram Mathematica<\/strong><\/em><\/a>, I was pleasantly surprised when the product was launched and experimented quite a bit with it. The product has grown substantially since then and is backed by a huge amount of \u201ccurated knowledge\u201d<\/strong><\/em>. What I like most about it is the way it presents its answers. Here is a sample:<\/p>\n

\"WolframAlpha<\/a>
WolframAlpha Query-1<\/strong><\/figcaption><\/figure>\n

It is remarkable that when I introduce brackets to indicate precedence in calculations, it is able to handle that correctly:<\/p>\n

\"Another<\/a>
Slightly Modified Query<\/strong><\/figcaption><\/figure>\n

The query need not be restricted to Mathematics or even Science; it can handle several domains. Here is a different query:<\/p>\n

\"A<\/a>
A General Query<\/strong><\/figcaption><\/figure>\n

How does this differ from ChatGPT<\/strong><\/em>? Let us see the answers generated by ChatGPT<\/strong><\/em> for the earlier two Math queries:<\/p>\n

\"ChatGPT<\/a>
ChatGPT Math Queries<\/strong><\/figcaption><\/figure>\n

As you can see, ChatGPT<\/strong><\/em> is not able to differentiate between the two queries. While I am not going to compare the two products in detail, it is only fair to point out that even WolframAlpha<\/strong><\/em> will fail in some cases. Here is one such example:<\/p>\n

\"WolframAlpha<\/a>
WolframAlpha Failed Query<\/strong><\/figcaption><\/figure>\n

Let us see how ChatGPT<\/strong><\/em> handles this case:<\/p>\n

\"ChatGPT<\/a>
ChatGPT Response<\/strong><\/figcaption><\/figure>\n

That is the correct answer!<\/p>\n

To me, what is interesting and special about WolframAlpha<\/strong><\/em> is not the NLP<\/strong><\/em> part or its knowledge base, but the fact that the query is internally converted into Expressions<\/strong><\/em> and Concepts<\/strong><\/em> built into the Wolfram<\/strong> Language<\/strong><\/em> and executed at the symbolic level.<\/p>\n

In a recent article<\/strong><\/em><\/a>, Stephen Wolfram has compared ChatGPT<\/strong><\/em> and WolframAlpha<\/strong><\/em> and gives examples where ChatGPT<\/strong><\/em> generated incorrect answers, while WolframAlpha<\/strong><\/em> gives the correct answers. He goes on to express his view that ChatGPT<\/strong><\/em> could perhaps use WolframAlpha<\/strong><\/em> as a \u201csuper power\u201d<\/strong><\/em> back end to compute the correct responses.<\/p>\n

While this is certainly possible and will work in some (or even many) cases, it might not solve the general problem of generating the correct response for an arbitrary input. I believe that a workable and scalable solution lies in adopting a different approach.<\/p>\n

The following diagram shows WolframAlpha<\/strong><\/em> and ChatGPT<\/strong><\/em> as they exist today:<\/p>\n

\"ChatGPT<\/a>
ChatGPT and WolframAlpha Today<\/strong><\/figcaption><\/figure>\n

Each product takes the input and generates the result directly (the fact that WolframAlpha<\/strong><\/em> converts this to a symbolic query internally is an implementation detail). This is a \u201cheavyweight\u201d<\/strong><\/em> approach where each tries to “do all”<\/strong><\/em> and “be all”<\/strong><\/em>. Not surprisingly, this leads to failures in some cases. It is impractical to be a \u201cmaster-of-everything\u201d<\/strong><\/em>! And this doesn\u2019t scale easily.<\/p>\n

Is there a different approach? I believe that the following scheme provides a better approach:<\/p>\n

\"A<\/a>
A Different Model<\/strong><\/figcaption><\/figure>\n

In this approach, we have a \u201cconverter\u201d<\/strong><\/em> that converts the free-form query to a formal meaning representation language, and this is then processed by a suitable domain expert to generate the correct response.<\/p>\n

What is this \u201cmeaning representation language\u201d<\/strong><\/em>? From Roger Schank\u2019s Conceptual Dependency Theory<\/strong> <\/em><\/a>to the more recent Abstract Meaning Representation Language (AMR)<\/strong><\/em><\/a>, a lot of interesting research is happening in the area of Meaning Representation<\/strong><\/em>. Of course, much more remains to be done.<\/p>\n

Here is what I would like to suggest: Just like we have a Java Virtual Machine (JVM)<\/em><\/strong><\/a>, we should devise a Cognitive Virtual Machine (CVM)<\/strong><\/em> architecture that mimics human cognition and, as part of that, define the primitive instructions that can be used in the Meaning Representation Language<\/strong><\/em>. This will help standardize the MRL.<\/strong><\/em> Domain experts can use this MRL<\/strong><\/em> to build custom intelligent solutions for their respective domains. I know this is a challenging problem to solve, but if the kind of collective effort (and money) that is being pumped into Machine Learning<\/strong><\/em> is also put into this CVM<\/strong><\/em> design, it will become a reality in the next decade.<\/p>\n

Building an intelligent system eventually becomes a \u201cplug-and-play\u201d<\/strong><\/em> system:<\/p>\n

\"Building<\/a>
Building Intelligent Solutions of the Future<\/strong><\/figcaption><\/figure>\n

An intelligent “chat agent”<\/strong><\/em> of the future can then be built by suitably integrating these components like this:<\/p>\n

\"Intelligent<\/a>
Intelligent Chat Agent of the Future<\/strong><\/figcaption><\/figure>\n

It might appear to be far-fetched today, but when experts from different disciplines come together, the Cognitive Virtual Machine<\/strong> <\/em>will be a reality and Artificial Intelligence<\/strong><\/em> will get a major boost at that time!<\/p>\n

Have a nice weekend!<\/p>\n","protected":false},"excerpt":{"rendered":"

We all know that ChatGPT has taken the world by storm. True, it is a major advance of Artificial Intelligence in the area of Natural Language Processing. Many may not know that WolframAlpha, launched in 2009, allowed natural language queries. As a long time user of Wolfram Mathematica, I was pleasantly surprised when the product […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"advanced_seo_description":"","jetpack_seo_html_title":"","jetpack_seo_noindex":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"jetpack_post_was_ever_published":false},"categories":[162,72,107],"tags":[348,351,350,349],"class_list":["post-2994","post","type-post","status-publish","format-standard","hentry","category-knowledge-representation","category-mathematica","category-natural-language-processing","tag-chatgpt","tag-cognitive-virtual-machine","tag-meaning-representation-language","tag-wolframalpha"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/p9OLnF-Mi","jetpack-related-posts":[{"id":328,"url":"https:\/\/www.rangakrish.com\/index.php\/2016\/09\/11\/natural-language-processing-in-mathematica\/","url_meta":{"origin":2994,"position":0},"title":"Natural Language Processing in Mathematica","author":"admin","date":"September 11, 2016","format":false,"excerpt":"Welcome back. Today I am going to share with you some of the nice capabilities of Mathematica in the area of Natural Language Processing (NLP). Let us start with words. What if we wish to know\u00a0the various definitions of the word image?\u00a0Here is the answer. Mathematica gives the various senses\u2026","rel":"","context":"In "Mathematica"","block_context":{"text":"Mathematica","link":"https:\/\/www.rangakrish.com\/index.php\/category\/mathematica\/"},"img":{"alt_text":"Word Definition","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/09\/word-data1-1024x238.png?resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/09\/word-data1-1024x238.png?resize=350%2C200 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/09\/word-data1-1024x238.png?resize=525%2C300 1.5x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/09\/word-data1-1024x238.png?resize=700%2C400 2x"},"classes":[]},{"id":2131,"url":"https:\/\/www.rangakrish.com\/index.php\/2020\/09\/13\/mathematica-using-textcases-to-extract-information-from-natural-language-text\/","url_meta":{"origin":2994,"position":1},"title":"Mathematica: Using TextCases to Extract Information from Natural Language Text\u00a0","author":"admin","date":"September 13, 2020","format":false,"excerpt":"Extracting meaningful information from unstructured, human readable text is a hot topic of research today and has important applications in many domains. I have written a few blogs related to this topic, for example, see this\u00a0and this. In today\u2019s article, I would like to show how Mathematica can be a\u2026","rel":"","context":"In "Mathematica"","block_context":{"text":"Mathematica","link":"https:\/\/www.rangakrish.com\/index.php\/category\/mathematica\/"},"img":{"alt_text":"Extracting Sentences","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2020\/09\/ex1-2-300x106.png?resize=350%2C200&ssl=1","width":350,"height":200},"classes":[]},{"id":1541,"url":"https:\/\/www.rangakrish.com\/index.php\/2019\/04\/21\/textcontents-function-in-mathematica-12\/","url_meta":{"origin":2994,"position":2},"title":"TextContents[ ] Function in Mathematica 12","author":"admin","date":"April 21, 2019","format":false,"excerpt":"Mathematica 12 was released a few days ago.\u00a0 It has been over a year since version 11.3 came out in March 2018. The long wait appears justified since the new release boasts of numerous improvements and new features across several areas. You may want to read this blog post\u00a0by Stephen\u2026","rel":"","context":"In "Mathematica"","block_context":{"text":"Mathematica","link":"https:\/\/www.rangakrish.com\/index.php\/category\/mathematica\/"},"img":{"alt_text":"Importing Text File","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2019\/04\/FileImport.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2019\/04\/FileImport.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2019\/04\/FileImport.png?resize=525%2C300&ssl=1 1.5x"},"classes":[]},{"id":870,"url":"https:\/\/www.rangakrish.com\/index.php\/2018\/03\/25\/question-answering-in-mathematica\/","url_meta":{"origin":2994,"position":3},"title":"Question Answering in Mathematica","author":"admin","date":"March 25, 2018","format":false,"excerpt":"About 10 days ago, I received an update for Mathematica. The latest version is 11.3.0. As usual, I looked through the list of new features\u00a0in this release. There are several new features, but one of them attracted my attention immediately: There is a new function FindTextualAnswer\u00a0that, given a piece of\u2026","rel":"","context":"In "Mathematica"","block_context":{"text":"Mathematica","link":"https:\/\/www.rangakrish.com\/index.php\/category\/mathematica\/"},"img":{"alt_text":"Example 1","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2018\/03\/Example1.png?resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2018\/03\/Example1.png?resize=350%2C200 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2018\/03\/Example1.png?resize=525%2C300 1.5x"},"classes":[]},{"id":3545,"url":"https:\/\/www.rangakrish.com\/index.php\/2024\/11\/09\/semantic-search-in-wolfram-mathematica\/","url_meta":{"origin":2994,"position":4},"title":"Semantic Search in Wolfram Mathematica","author":"admin","date":"November 9, 2024","format":false,"excerpt":"In an earlier article, I explained how to use OpenAI from Wolfram Mathematica ver 14.1. This latest release of Wolfram supports Semantic Search as well. In today\u2019s article, let me discuss this feature. As in the case of using LLMs, using Semantic Search requires an account with one of the\u2026","rel":"","context":"In "Mathematica"","block_context":{"text":"Mathematica","link":"https:\/\/www.rangakrish.com\/index.php\/category\/mathematica\/"},"img":{"alt_text":"Remedy Description","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2024\/11\/image1-300x225.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2024\/11\/image1-300x225.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2024\/11\/image1-300x225.png?resize=525%2C300&ssl=1 1.5x"},"classes":[]},{"id":409,"url":"https:\/\/www.rangakrish.com\/index.php\/2016\/11\/02\/working-with-linguistic-data-in-mathematica\/","url_meta":{"origin":2994,"position":5},"title":"Working with Linguistic Data in Mathematica","author":"admin","date":"November 2, 2016","format":false,"excerpt":"There are many interesting functions in Mathematica for working with language data, not just in English but in many other languages too. The DictionaryLookup[] function is a good starting point. Let us see what languages are supported as part of dictionary lookup: That is a good collection. It is nice\u2026","rel":"","context":"In "Mathematica"","block_context":{"text":"Mathematica","link":"https:\/\/www.rangakrish.com\/index.php\/category\/mathematica\/"},"img":{"alt_text":"Supported Languages","src":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/11\/dict-1.png?resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/11\/dict-1.png?resize=350%2C200 1x, https:\/\/i0.wp.com\/www.rangakrish.com\/wp-content\/uploads\/2016\/11\/dict-1.png?resize=525%2C300 1.5x"},"classes":[]}],"_links":{"self":[{"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/posts\/2994","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/comments?post=2994"}],"version-history":[{"count":0,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/posts\/2994\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/media?parent=2994"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/categories?post=2994"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.rangakrish.com\/index.php\/wp-json\/wp\/v2\/tags?post=2994"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}