{"id":794,"date":"2024-04-12T16:12:20","date_gmt":"2024-04-12T22:12:20","guid":{"rendered":"https:\/\/redmonk.com\/rstephens\/?p=794"},"modified":"2024-04-15T08:52:14","modified_gmt":"2024-04-15T14:52:14","slug":"monkigras-2","status":"publish","type":"post","link":"https:\/\/redmonk.com\/rstephens\/2024\/04\/12\/monkigras-2\/","title":{"rendered":"Monki Gras Recap, Day 2"},"content":{"rendered":"<p>We\u2019re recapping the talks presented at the 2024 Monki Gras. Below is a summary of the talks from the final day of the conference. <a href=\"https:\/\/redmonk.com\/rstephens\/2024\/04\/05\/monkigras-1\/\">A summary of the talks from the first day is available here.<\/a><\/p>\n<hr \/>\n<h2>Stochastic Children by Dormain Drewitz and Rachel Stephens<\/h2>\n<p>Drewitz and Stephens (oh hai!) explored the overlap between effectively prompting as parents and prompting an LLM. Can we think like parents in how we approach our interactions with LLMs, and can LLMs show us anything about being better parents?<\/p>\n<p>We explored the ways we get our children to do things and whether it translates to LLMs. We also highlighted the tradeoffs that come from spending the time (sometimes years!) to train the model\/child with necessary context so your prompts can be brief and one-shot, as compared to providing in-depth instructions in the moment.<\/p>\n<p>Also: the power of sticker charts!<\/p>\n<p><a href=\"https:\/\/twitter.com\/taylor_atx\/status\/1768594196607406539\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.52.22@2x-520x620.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"620\" class=\"aligncenter size-medium wp-image-795\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.52.22@2x-520x620.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.52.22@2x-859x1024.png 859w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.52.22@2x-768x916.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.52.22@2x-480x572.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.52.22@2x-526x627.png 526w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.52.22@2x.png 988w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<h2>A week\u2019s work in 6 hours \u2013 how LLMs will revolutionise the study of humanities by Ian Miell<\/h2>\n<p>Miell used LLMs to rewrite one of his essays from university without referencing the original essay and without re-reading the texts. The prompt was \u2018How was new Liberalism different from the old Liberalism?\u2019<\/p>\n<p>Miell spoke about the power and limitations of LLMs. While there were definitely limitations in the LLMs ability to do deeper thinking, it was an excellent tool reducer and learning accelerator. It can\u2019t write a decent university-level history essay yet, but it shone at helping plan the essay writing.<\/p>\n<p>He also included such insightful references to Douglas Adams and Garry Kasparov.<\/p>\n<p><a href=\"https:\/\/twitter.com\/patrickdebois\/status\/1768596197189538088\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.53.57@2x-520x544.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"544\" class=\"aligncenter size-medium wp-image-796\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.53.57@2x-520x544.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.53.57@2x-978x1024.png 978w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.53.57@2x-768x804.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.53.57@2x-480x502.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.53.57@2x-599x627.png 599w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.53.57@2x.png 984w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p><a href=\"https:\/\/twitter.com\/sogrady\/status\/1768658040746586485\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.55.08@2x-520x539.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"539\" class=\"aligncenter size-medium wp-image-797\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.55.08@2x-520x539.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.55.08@2x-988x1024.png 988w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.55.08@2x-768x796.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.55.08@2x-480x497.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.55.08@2x-605x627.png 605w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.55.08@2x.png 992w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p>Miell\u2019s <a href=\"https:\/\/docs.google.com\/presentation\/d\/1vSljYGu-bcd1d93FYKblnw8_7CoBpGIcf9VUcozYCWM\/edit#slide=id.g742e3e7cd_1_16\">slides are available here<\/a>.<\/p>\n<h2>Cake: an open source tale by Julia Ferraioli<\/h2>\n<p>Ferraioli shared the primary tenants of open source:<\/p>\n<ul>\n<li>can I learn about the code and examine its inputs and outputs?<\/li>\n<li>can I inspect it and find bugs?<\/li>\n<li>can I change it to suit my purposes?<\/li>\n<li>can I give it to others? <\/li>\n<\/ul>\n<p>People have expectations of what they can do when something is called \u2018open source,\u2019 and the LLM community is not always meeting these expectations with their current language.<\/p>\n<p>The problem we sometimes face with LLMs is that 1. people use \u2018open source\u2019 inconsistently and 2. it\u2019s not always clear how to apply some of the core principles of open source to key LLM concepts (for example, what does it mean to share how the input data was cleaned in an open source way?) This is a challenging problem, but one we need to pay attention to.<\/p>\n<p><a href=\"https:\/\/twitter.com\/taylor_atx\/status\/1768603284867010887\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.05@2x-520x567.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"567\" class=\"aligncenter size-medium wp-image-798\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.05@2x-520x567.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.05@2x-940x1024.png 940w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.05@2x-768x837.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.05@2x-480x523.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.05@2x-575x627.png 575w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.05@2x.png 982w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<h2>Lines in the sand by Jim Boulton<\/h2>\n<p>Boulton is a tech historian who tells stories about unsung tech heros using comics. Boulton shared his experience creating a comic book about <a href=\"https:\/\/en.wikipedia.org\/wiki\/Lynn_Conway\">Lynn Conway<\/a> using LLM prompts to generate the images. Each image required hundreds of iterations to create a cohesive look and feel throughout the comic.<\/p>\n<p><a href=\"https:\/\/twitter.com\/patrickdebois\/status\/1768607827424104863\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.59@2x-520x521.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"521\" class=\"aligncenter size-medium wp-image-799\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.59@2x-520x521.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.59@2x-150x150.png 150w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.59@2x-768x770.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.59@2x-480x481.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.59@2x-300x300.png 300w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.59@2x-626x627.png 626w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.56.59@2x.png 974w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p>I love the above quote, but Boulton\u2019s talk also had one of my favorite lines from the event: \u201cI discovered AI\u2019s not very good at fingers, but then I talked to an illustrator and they said \u2018illustrators aren\u2019t really good at fingers, either.\u2019\u201d<\/p>\n<p>His comics on Lynn Conway and Alan Emtage are delightful and can be <a href=\"https:\/\/www.unsungheroes.info\/catalog\">purchased here<\/a>.<\/p>\n<p><a href=\"https:\/\/twitter.com\/jim_boulton\/status\/1768732054751985719\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.57.46@2x-520x476.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"476\" class=\"aligncenter size-medium wp-image-800\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.57.46@2x-520x476.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.57.46@2x-768x703.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.57.46@2x-480x439.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.57.46@2x-685x627.png 685w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.57.46@2x.png 990w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<h2>\u201cI Didn\u2019t Grow Up Speaking Code\u201d: GitHub Copilot as a Programming-as-a-Second-Language Tool by Kristen Foster-Marks<\/h2>\n<p>Foster-Marks\u2019 talk was a delightful dive into what it means to master coding as a language. Her talk drew parallels to what second language acquisition; it\u2019s not enough to be able to code, but we also have to be able to \u201cspeak code\u201d to describe what it\u2019s doing. What are some examples of when we \u201cspeak code\u201d?<\/p>\n<ul>\n<li>interviews<\/li>\n<li>PRs and code reviews<\/li>\n<li>paired programming<\/li>\n<li>live streaming<\/li>\n<li>pseudo coding<\/li>\n<li>explaining things to business stakeholders<\/li>\n<li>rubber ducking<\/li>\n<\/ul>\n<p>Foster-Marks then explored the consequences of poor \u201cfluency\u201d for both individuals and teams, and then discussed how AI code generation tools like GitHub Copilot can help enhance this skill.<\/p>\n<p><a href=\"https:\/\/www.kristen-foster-marks.com\/post\/i-didnt-grow-up-speaking-code\">Read more about Foster-Mark\u2019s talk here<\/a><\/p>\n<p><a href=\"https:\/\/twitter.com\/DormainDrewitz\/status\/1768621926300127667\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.58.38@2x-520x641.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"641\" class=\"aligncenter size-medium wp-image-801\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.58.38@2x-520x641.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.58.38@2x-831x1024.png 831w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.58.38@2x-768x946.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.58.38@2x-480x592.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.58.38@2x-509x627.png 509w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.58.38@2x.png 990w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<h2>From Text to Flaws: vulnerabilities in applications with Generative AI and LLMs by Paul Molin<\/h2>\n<p>Molin spoke about prompt injections and how to ameliorate their harms. As the popularity of GenAI grows, we need to understand how it can be exploited and what options we have to protect against manipulation or data exfiltration.<\/p>\n<p><a href=\"https:\/\/twitter.com\/DormainDrewitz\/status\/1768629957968748720\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.59.36@2x-520x546.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"546\" class=\"aligncenter size-medium wp-image-802\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.59.36@2x-520x546.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.59.36@2x-975x1024.png 975w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.59.36@2x-768x807.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.59.36@2x-480x504.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.59.36@2x-597x627.png 597w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-15.59.36@2x.png 988w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p>Some potential steps to limit vulnerabilities include:<\/p>\n<ol>\n<li>Use LLM to analyze input\/output<\/li>\n<li>Compare the input to a list of known injections <\/li>\n<li>Use 2 LLMs with different accesses<\/li>\n<li>Preflight prompts, detect injections before execution <\/li>\n<li>Human in the loop<\/li>\n<li>Canaries<\/li>\n<\/ol>\n<p><a href=\"https:\/\/twitter.com\/DormainDrewitz\/status\/1768632827774398574\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.00.36@2x-520x767.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"767\" class=\"aligncenter size-medium wp-image-803\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.00.36@2x-520x767.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.00.36@2x-694x1024.png 694w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.00.36@2x-768x1133.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.00.36@2x-480x708.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.00.36@2x-425x627.png 425w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.00.36@2x.png 994w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p>(That said, my favorite part of the talk was the intro where Molin compared the sizes of cats and bananas, which appealed to one of my favorite things, measuring using uncommon metrics (like <a href=\"https:\/\/youtu.be\/4b4TESNlZUo?si=j-33jyjzO-uXUGmz\">giraffes\/hour<\/a>)<\/p>\n<p><a href=\"https:\/\/twitter.com\/DormainDrewitz\/status\/1768627384939352460\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.01.26@2x-520x500.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"500\" class=\"aligncenter size-medium wp-image-804\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.01.26@2x-520x500.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.01.26@2x-768x739.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.01.26@2x-480x462.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.01.26@2x-652x627.png 652w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.01.26@2x.png 992w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<h2>Working with ChatGPT by Jessica West<\/h2>\n<p>West shared some behind the scenes stories about creating the visual assets used for this year\u2019s Monki Gras. West collaborated with designer Jack James to create the logos and imagery for the event, but\u2013given the theme of the conference\u2013the team of course used GenAI as an element of the design process.<\/p>\n<p>West walked through the design team\u2019s iterative process of brainstorming and creating with ChatGPT, and detailed the challenges of trying to bring the GenAI concepts to fruition with James\u2019 design work.<\/p>\n<p>(Also, be sure to see her infographics with pictures of James Governor as \u2018beer Jesus.\u2019)<\/p>\n<p><a href=\"https:\/\/twitter.com\/DormainDrewitz\/status\/1768649859232387313\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.02.09@2x-520x475.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"475\" class=\"aligncenter size-medium wp-image-805\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.02.09@2x-520x475.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.02.09@2x-768x701.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.02.09@2x-480x438.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.02.09@2x-687x627.png 687w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.02.09@2x.png 992w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<h2>Why prompt engineering is BS by Kyle Roche<\/h2>\n<p>Griptape, co-founded by Roche, focuses on creating \u201coff-prompt\u201d AI solutions. Their product hypothesis centers around developer tooling preferences and enterprise data privacy concerns, and thus they created a framework and opinionated deployment environment for writing AI-powered Python applications. Roche\u2019s talk discussed a vision when AI developers don\u2019t have to care about prompt engineering at all.<\/p>\n<p><a href=\"https:\/\/twitter.com\/patrickdebois\/status\/1768651797701308597\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.08@2x-520x523.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"523\" class=\"aligncenter size-medium wp-image-806\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.08@2x-520x523.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.08@2x-150x150.png 150w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.08@2x-768x773.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.08@2x-480x483.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.08@2x-623x627.png 623w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.08@2x.png 984w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<h2>Generative AI\u2019s Hallucination Problem and How to Fix It. by Emil Eifrem<\/h2>\n<p>Eifrem\u2019s talk did a fantastic job connecting related technology concepts, as talked about his experience with graph technology as a co-founder of Neo4j related to this next generation of AI tools. He explained knowledge graphs and how they relate to retrieval augmented generation (RAG).<\/p>\n<p><a href=\"https:\/\/twitter.com\/lilytownsend23\/status\/1768656138264883372\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.48@2x-520x478.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"478\" class=\"aligncenter size-medium wp-image-807\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.48@2x-520x478.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.48@2x-768x706.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.48@2x-480x441.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.48@2x-682x627.png 682w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.03.48@2x.png 984w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p>Eifrem called AI a once in a generation shift. While RAG can add context to an LLM, GraphRAG can unlock additional context from the knowledge graph and help LLMs avoid hallucinations.<\/p>\n<p><a href=\"https:\/\/twitter.com\/DormainDrewitz\/status\/1768657679570862542\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.04.37@2x-520x570.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"570\" class=\"aligncenter size-medium wp-image-808\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.04.37@2x-520x570.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.04.37@2x-933x1024.png 933w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.04.37@2x-768x842.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.04.37@2x-480x527.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.04.37@2x-572x627.png 572w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.04.37@2x.png 990w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p><a href=\"https:\/\/twitter.com\/monkigras\/status\/1767586414026186764\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.05.24@2x-520x471.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"471\" class=\"aligncenter size-medium wp-image-809\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.05.24@2x-520x471.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.05.24@2x-768x695.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.05.24@2x-480x434.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.05.24@2x-693x627.png 693w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.05.24@2x.png 988w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<h2>AI for Developer Productivity by Paige Bailey<\/h2>\n<p><a href=\"https:\/\/twitter.com\/sogrady\/status\/1768661852127482342\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.05@2x-520x694.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"694\" class=\"aligncenter size-medium wp-image-810\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.05@2x-520x694.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.05@2x-767x1024.png 767w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.05@2x-768x1026.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.05@2x-480x641.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.05@2x-470x627.png 470w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.05@2x.png 978w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p>Bailey gave us a behind the scenes tour of how Google is using language models and copilots to build internally. She discussed how GenAI can reduce the friction from having an idea to implementing it, with the goal of \u201cgiving developers more good days.\u201d<\/p>\n<p><a href=\"https:\/\/twitter.com\/DormainDrewitz\/status\/1768662336498212996\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.49@2x-520x622.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"622\" class=\"aligncenter size-medium wp-image-811\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.49@2x-520x622.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.49@2x-856x1024.png 856w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.49@2x-768x918.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.49@2x-480x574.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.49@2x-524x627.png 524w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.06.49@2x.png 980w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<p>In particular I liked her examples of advances are now happening in days and weeks, not decades. I also liked her context around what 1-10M tokens actually \u201cfeels\u201d like.<\/p>\n<p><a href=\"https:\/\/twitter.com\/DormainDrewitz\/status\/1768663951154319825\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.07.34@2x-520x564.png\" alt=\"Tweet screenshot (click link for full link to tweet)\" width=\"520\" height=\"564\" class=\"aligncenter size-medium wp-image-812\" srcset=\"https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.07.34@2x-520x564.png 520w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.07.34@2x-944x1024.png 944w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.07.34@2x-768x833.png 768w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.07.34@2x-480x521.png 480w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.07.34@2x-578x627.png 578w, https:\/\/redmonk.com\/rstephens\/files\/2024\/04\/CleanShot-2024-04-12-at-16.07.34@2x.png 990w\" sizes=\"auto, (max-width: 520px) 100vw, 520px\" \/><\/a><\/p>\n<hr \/>\n<p>Thanks again to all speakers, sponsors, and attendees. You made the return of Monki Gras special!<\/p>\n","protected":false},"excerpt":{"rendered":"<p>We\u2019re recapping the talks presented at the 2024 Monki Gras. Below is a summary of the talks from the final day of the conference. A summary of the talks from the first day is available here. Stochastic Children by Dormain Drewitz and Rachel Stephens Drewitz and Stephens (oh hai!) explored the overlap between effectively prompting<\/p>\n","protected":false},"author":45,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"spay_email":"","footnotes":"","jetpack_publicize_message":"Monki Gras Recap, Day 2","jetpack_is_tweetstorm":false},"categories":[40,42,133],"tags":[],"class_list":["post-794","post","type-post","status-publish","format-standard","hentry","category-ai","category-conferences","category-monki-gras"],"jetpack_featured_media_url":"","jetpack_publicize_connections":[],"jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/posts\/794","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/users\/45"}],"replies":[{"embeddable":true,"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/comments?post=794"}],"version-history":[{"count":0,"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/posts\/794\/revisions"}],"wp:attachment":[{"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/media?parent=794"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/categories?post=794"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/redmonk.com\/rstephens\/wp-json\/wp\/v2\/tags?post=794"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}