{"id":607232,"date":"2023-02-12T07:49:55","date_gmt":"2023-02-12T13:49:55","guid":{"rendered":"https:\/\/news.sellorbuyhomefast.com\/index.php\/2023\/02\/12\/the-language-revolution-how-llms-could-transform-the-world\/"},"modified":"2023-02-12T07:49:55","modified_gmt":"2023-02-12T13:49:55","slug":"the-language-revolution-how-llms-could-transform-the-world","status":"publish","type":"post","link":"https:\/\/newsycanuse.com\/index.php\/2023\/02\/12\/the-language-revolution-how-llms-could-transform-the-world\/","title":{"rendered":"The language revolution: How LLMs could transform the world"},"content":{"rendered":"<div>\n<section>\n<p><time title=\"2023-02-11T16:20:00+00:00\" datetime=\"2023-02-11T16:20:00+00:00\">February 11, 2023 8:20 AM<\/time>\n\t\t\t<\/p>\n<\/section>\n<div>\n<p><img loading=\"lazy\" decoding=\"async\" width=\"750\" height=\"513\" src=\"https:\/\/venturebeat.com\/wp-content\/uploads\/2022\/02\/GettyImages-1367728606-e1645809619993.jpg?fit=750%2C513&#038;strip=all\" alt=\"Conversational AI Concept - Natural Language Processing - NLP - Computational Linguistics Concept - AI-based Virtual Assistant Generating Voice as a Sound Wave\"><\/p>\n<div>\n<p><em>Image Credit: Getty Images<\/em><\/p>\n<\/div><\/div>\n<\/p><\/div>\n<div id=\"primary\" role=\"main\">\n<article id=\"post-2845930\">\n<div>\n<div id=\"boilerplate_2682874\">\n<p><em>Check out all the on-demand sessions from the Intelligent Security Summit <a href=\"https:\/\/avolio.swapcard.com\/intelligentsecuritysummit2022\/registrations\/Start?utm_source=vb&#038;utm_medium=boiler&#038;utm_content=ondemand&#038;utm_campaign=IS22_BoilerPlates\" data-type=\"URL\" data-id=\"https:\/\/avolio.swapcard.com\/intelligentsecuritysummit2022\/registrations\/Start?utm_source=vb&#038;utm_medium=boiler&#038;utm_content=ondemand&#038;utm_campaign=IS22_BoilerPlates\">here<\/a><\/em>.<\/p>\n<hr>\n<\/div>\n<p>We are living at a historic moment. A new revolution, comparable to the Industrial Revolution, is underway. Entire industries are going to be disrupted. The nature of creativity and knowledge work is going to change. Language is going to become the most important sense for humans. Language \u2014 specifically in the form of large language models (LLMs) \u2014 is going to reshape how we think about the world around us.<\/p>\n<p>Every once in a while, technology reaches an inflection point that leads to a paradigm shift. That\u2019s what\u2019s happening now, and we\u2019re just at the beginning. <a href=\"https:\/\/venturebeat.com\/ai\/how-mit-is-training-ai-language-models-in-an-era-of-quality-data-scarcity\/\">LLMs<\/a> like GPT-3, are getting really good at generating text, summarizing text, reasoning, understanding, writing poetry and more. They are the world\u2019s best autocomplete. They are changing how people write code, poems, marketing copies, essays, research papers, and more. They are not replacing jobs, but augmenting them, making us more productive.<\/p>\n<p>Of course, LLMs are far from perfect and have many challenges, such as hallucination, alignment and truthfulness. These are hard problems to solve, but solving them will make these models and applications much more reliable and robust.<\/p>\n<h2 id=\"h-sparking-the-rise-of-llms\">Sparking the rise of LLMs<\/h2>\n<p><a href=\"https:\/\/venturebeat.com\/ai\/how-chatgpt-in-microsoft-office-could-change-the-workplace-the-ai-beat\/\">ChatGPT<\/a> was the spark that ignited this fire. It showed how things got real when it went from zero to one million users in four days. Silicon Valley has started to build great applications and companies on top of LLMs, laying the foundation for the next trillion-dollar-valuation companies. We\u2019re also seeing the birth of new industries that are built with automation first, and human-in-the-loop second. These are what I call <a href=\"https:\/\/venturebeat.com\/2022\/06\/15\/what-is-artificial-intelligence\/\">AI<\/a>-first companies.<\/p>\n<p><html><body><\/p>\n<div id=\"boilerplate_2803147\">\n<h3>Event<\/h3>\n<div>\n<p><span>Intelligent Security Summit On-Demand<\/span><\/p>\n<p><span>Learn the critical role of AI &#038; ML in cybersecurity and industry specific case studies. Watch on-demand sessions today.<\/span><\/p>\n<\/div>\n<p><a href=\"https:\/\/avolio.swapcard.com\/intelligentsecuritysummit2022\/registrations\/Start?utm_source=vb&#038;utm_medium=incontent&#038;utm_content=ondemand&#038;utm_campaign=IS22_InContent\"><br \/>\n                Watch Here            <\/a>\n                        <\/p>\n<\/div>\n<p><\/body><\/p>\n<p>One of the great joys in life is experiencing art that resonates with us on an emotional level. As <a href=\"https:\/\/venturebeat.com\/2022\/06\/17\/what-is-generative-artificial-intelligence-ai\/\">generative AI<\/a> advances, I look forward to the ways it will enable us to tap into our creative potential even more, democratizing the process of authentic self-expression.<\/p>\n<p>But how do you build a moat around this? How do you capture value? To my mind, the key moats for LLM\/AI-first applications, in order of importance, are:<\/p>\n<ul>\n<li>Proprietary data and fine-tuning<\/li>\n<li>Great UX, and one that instills a sense of trust and reliability<\/li>\n<li>Cost to serve\/operationalize<\/li>\n<li>Distribution and GTM<\/li>\n<li>Network effects and community<\/li>\n<li>Breadth and depth of integrations<\/li>\n<\/ul>\n<p>Here\u2019s what I mean by the breadth and depth of integrations: Thin layers around LLM <a href=\"https:\/\/venturebeat.com\/2022\/06\/11\/apis-create-digital-empathy\/\">APIs<\/a> are not enough to gain a competitive edge in AI-first apps. To win, you need deep integrations and optimized workflows that solve real problems with the scalability and efficiency that wasn\u2019t possible before LLMs. For example, imagine using LLMs to augment teachers to create exam questions for students by:<\/p>\n<ul>\n<li>Providing a link to the content material<\/li>\n<li>Fetching\/scraping the content and parsing it into a format that LLMs understand better<\/li>\n<li>Asking LLMs to create questions from that content, given preferences like difficulty, etc.<\/li>\n<li>Using LLMs to write truthful answers to the questions from the content<\/li>\n<li>Using the edited answers to improve the future generations of questions<\/li>\n<\/ul>\n<p>This is just one example, but there are many verticals I\u2019m excited about: end-to-end SDR automation, code generation and refactoring, customer support automation, script-writing, medical\/health assistants, and education. AI-first apps will transform how we work and collaborate over the next five years, making knowledge work and intelligence more accessible and affordable. Note-taking and copyrighting are just the tip of the iceberg. New interfaces, CRMs, tax prep copilots, research assistants are all fair game.<\/p>\n<h2>LLMs now and in the future<\/h2>\n<p>Here\u2019s how I see the stages of LLM development:<\/p>\n<ul>\n<li>1.0: Capable of generating original text and reasoning about it<\/li>\n<li>2.0: Able to evolve, refine its output, and acquire new abilities to act rationally<\/li>\n<li>3.0: Can design its own actions\/capabilities to interact with the external world<\/li>\n<li>4.0+: Leverages the data flywheel to improve over time, and maintains itself<\/li>\n<\/ul>\n<p>The LLM landscape is increasingly starting to look something like this:<\/p>\n<ul>\n<li>Model layer (e.g. GPT-3, Cohere)<\/li>\n<li>API bindings for access (e.g. OpenAI Python)<\/li>\n<li>Infra layer for prompt chaining\/model switching (e.g. LangChain, Humanloop)<\/li>\n<li>Next-gen AI-first apps<\/li>\n<\/ul>\n<p>Within the infra layer, there are a few areas I find increasingly interesting: tooling\/infra, <a href=\"https:\/\/venturebeat.com\/dev\/ai-and-low-no-code-what-they-can-and-cant-do-together\/\">no\/low code<\/a>, fine-tuning, prompt chaining and retrieval, actions, experimentation frameworks. Creating a reliable and adaptable layer of infrastructure and tools for LLMs will help us unlock their power and value for more users and applications. To be honest, the recursive richness of LLM prompt chaining will revolutionize whole industries. (Or maybe I just find recursive things particularly fascinating.)<\/p>\n<p>Moreover, I agree that the next generation of AI-native products will integrate some elements of combining reasoning and acting in LLMs to help with decision-making. I like how Denny Zhou puts it: \u201cIf LLMs are humans, all the ideas are trivial: chain-of-thought prompting (\u2018explain your answer\u2019), self-consistency (\u2018double check your answer\u2019), least-to-most prompting (\u2018decompose to easy subproblems\u2019). The shocking thing is that LLMs are not humans but these still work!\u201d<\/p>\n<p>So, let\u2019s embrace the opportunity to work alongside intelligent systems that can help us unlock our full potential. The best platforms powered by LLMs will revolve around collaborative environments where humans and AI can work together. Together, we can achieve more than we ever thought possible.<\/p>\n<p><em>Shyamal Hitesh Anadkat works in applied AI at OpenAI<\/em>.<\/p>\n<div id=\"boilerplate_2736392\">\n<h3 id=\"h-datadecisionmakers\">DataDecisionMakers<\/h3>\n<p>Welcome to the VentureBeat community!<\/p>\n<p>DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.<\/p>\n<p>If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.<\/p>\n<p>You might even consider\u00a0<a rel=\"noreferrer noopener\" target=\"_blank\" href=\"https:\/\/venturebeat.com\/contribute-to-datadecisionmakers\/\">contributing an article<\/a>\u00a0of your own!<\/p>\n<p><a rel=\"noreferrer noopener\" href=\"https:\/\/venturebeat.com\/category\/DataDecisionMakers\/\" target=\"_blank\">Read More From DataDecisionMakers<\/a><\/p>\n<\/div>\n<p>\t\t\t\t<\/html><\/div>\n<\/p><\/div>\n<p><a href=\"https:\/\/venturebeat.com\/ai\/the-language-revolution-how-llms-could-transform-the-world\/\" class=\"button purchase\" rel=\"nofollow noopener\" target=\"_blank\">Read More<\/a><br \/>\n Shyamal Anadkat, OpenAI<\/p>\n","protected":false},"excerpt":{"rendered":"<p>February 11, 2023 8:20 AM Image Credit: Getty Images Check out all the on-demand sessions from the Intelligent Security Summit here. We are living at a historic moment. A new revolution, comparable to the Industrial Revolution, is underway. Entire industries are going to be disrupted. The nature of creativity and knowledge work is going to<\/p>\n","protected":false},"author":1,"featured_media":607233,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[4105,3030,46],"tags":[],"class_list":{"0":"post-607232","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-language","8":"category-revolution","9":"category-technology"},"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/posts\/607232","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/comments?post=607232"}],"version-history":[{"count":0,"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/posts\/607232\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/media\/607233"}],"wp:attachment":[{"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/media?parent=607232"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/categories?post=607232"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/newsycanuse.com\/index.php\/wp-json\/wp\/v2\/tags?post=607232"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}