Crazy Wisdom

Show Notes for the Crazy Wisdom Podcast Guest: Jared Zhao, CEO of AskEdith (https://www.askedith.ai/) Conversation: Introduction to AskEdith: Explanation of the search-driven analytics tool that enables users to access data in plain language Targeted at non-technical users Background: Jared's previous work in creating a visual editor for a data warehouse and training non-technical users Evolution of AskEdith: The journey so far and how it's still in its early stages The influence of startup circles and the emergence of large language models (LLMs) like ChatGPT from OpenAI The potential for swapping out the underlying core technology The role of English to SQL translations Future of AskEdith and Industry Insights: Thoughts on the future of core IP and the impact of open source Understanding the inference structure and use of APIs Discussion on the current competitive advantage of OpenAI The litmus test for whether we are in a technological singularity and its definition Unique aspects of the data industry, including the inability to host data Technical Aspects of AskEdith: The balance between search functionality, security, and compliance The use of a terraform module for one-click deployment into private clouds Insights into database and networking configuration AI and Society: The implications of AI on philosophy and its potential as a "silver bullet" The difference between AI replacing tasks versus jobs The historical context of automation and the Roman Empire's approach to universal basic income The potential need for universal business income today Whether AI will lead to deflation and thoughts on the economics of AI AI in Practice: The role of AI in replacing resources like Stack Overflow Challenges faced by AI in writing innovative code and the current capabilities of GPT-4 The potential for AI to predict the future and become an expert in a field The use of external data and consumer trends for benchmarking The Future of AskEdith: The vision of AskEdith as a mission control for enterprises The role of AskEdith as a search layer

Introduction Jae Yang, the founder of Tacen, a decentralized exchange, discusses his first hire as an archivist for the company and how this unique approach to knowledge management has helped them navigate the complex world of crypto startups. Tacen's commitment to thorough documentation and keeping a permanent record of every decision made is a testament to their belief in transparency and accessibility. Topics Covered Knowledge Management and the Role of an Archivist Jae shares how his experience at various startups led him to the role of an archivist, a guardian of a company's history and knowledge, differentiating it from a librarian who curates and organizes knowledge. The importance of meeting notes and how AI is utilized in their decision support process. The evolution of their team, starting as the 3rd person on board, to now 26 members. The process of making knowledge visible across the company and its challenges. Explains their unique classification system for information security that ranges from '00' for public data to '04' for secret data. Tools Used for Knowledge Management The use of internal wikis and Urbit, a suite of software tools that provides good infrastructure level support. Transition from Notion to WikiJS for self-hosting, and the reason behind the move. The capabilities of Urbit in knowledge management, including indexing textual content and its limitations like the absence of a textual management system like Obsidian. Urbit vs Obsidian The comparison between Urbit and Obsidian, which has features like bi-directional tagging and infinite discoverability. Events Discussion of "Reassembly 2023," a conference hosted by Tacen in Cheyenne, Wyoming from August 16-18th. Company Location and Cryptocurrency Law Explanation of why Tacen was established in Wyoming, citing factors such as crypto-friendly laws like the Wyoming DAO Law, zero corporate tax, and good internet infrastructure. Comparisons to South Dakota's innovative banking regulations and credit card processing. Brazil and Cryptocurrency The situation of cryptocurrency in Brazil, including its integration in daily life with apps and its role in circumventing high inflation and tax burdens. Brief mention of cryptocurrency's role in Argentina (note: segment about stablecoins was cut out). World View and Social Change Jae's thoughts on the rapidly changing world, the social changes since the 1970s, and the impact of technology. Discussion on the falling birth rates, changing attitudes to personal rights, and the further atomization of society. His view of the internet as a powerful tool that can lead to both individual empowerment and the spread of harmful agendas. Banks and Access to Finance Concern over banks preventing people from accessing their accounts for ideological reasons. Cheyenne and Tech Startups The vibrant tech startup scene in Cheyenne and its nearby University town, Laramie. About Tacen Tacen's early struggles with information overload and automatic deletion of data on Notion, leading to a change in their documentation strategy.

Robert DT on twitter: @DeeperThrill Doctorate on biomedical engineering with a focus on AI Entrepreneur building biomedical systems with AI specifically; medical imaging The conversation centers on the role of artificial intelligence (AI) in medical imaging, with an emphasis on computer vision and the utilization of existing imaging algorithms. Transformers, a type of deep learning model, are discussed for their unique self-attention mechanism and applications in natural language processing and computer vision. The talk pivots to data cleaning, specifically anonymization and safeguarding personal identifiers in the context of healthcare. Questions arise about data storage in healthcare facilities and the process of transferring it to the cloud. The conversation broadens to encompass AI's predictive capabilities and inherent risks, including the possibility of AI predicting third-order effects of its own interventions and concerns about excessive trust in AI predictions. The potential of AI in genetic engineering surfaces, particularly regarding CRISPR technology and nanobots. The conversation explores the benefits and risks of such advancements, including the revival of extinct plants and emergence of new diseases. Finally, the conversation shifts to societal implications of AI, including job displacement, the emergence of an attention economy, and the prospects of decentralized AI. The importance of understanding the limits of AI is underscored.   Show notes We need to examine what's currently happening in the field of AI, particularly in relation to medical imaging. This involves an exploration of computer vision technologies and how pre-existing imaging algorithms are being applied. We should discuss the concept of a "transformer" in the context of artificial intelligence. A critical part of working with AI is data cleaning. This includes the process of anonymization, ensuring that we only use the person's image and not any identifiable data like their name. We must also consider the storage of this data, which is typically housed on hospital servers. Additionally, there's the question of how this data is transferred to a cloud system for further processing. Let's explore the issue of gatekeeping in the field of AI. This might involve discussing the role of clinical trials and the Institutional Review Board in ensuring ethical standards. The engineering aspect of gatekeeping also requires attention, particularly when dealing with 3D data sets for imaging. We should highlight two major changes currently happening in the field of AI. Swin Transformers represent a significant development, as they are built off the concept of transformers in AI. Let's delve into the world of language modeling and chatbots. We must also consider the potential downsides of these AI technologies. The transhumanism angle presents an interesting point of discussion, particularly in relation to the next generation of technology. For example, the development of the mRNA vaccine was a major leap forward in response to global health crises. There's also the concept of generative mRNA vaccines, which use AI to generate potential cures. However, these AI technologies also come with risks. They could inadvertently create a disease, or develop a cure that isn't effective. The ease with which technology can be used in this field means that virtually anyone can make implants, leading to a new set of challenges. We should also discuss the emerging role of AI in lab-based work, such as managing petri dishes. The application of Hegelian principles to AI provides an interesting philosophical perspective. Looking ahead, we might consider what a lab kit might look like in ten years. The idea of the first version of something, and its relationship to anti-authoritarianism, is another interesting topic to explore. We have to acknowledge that AI, despite its potential, will not prevent all risks. AI can be used as a predictive tool for triaging, helping to determine whether an intervention will benefit a person. The use of CRISPR technology is another relevant point of discussion, especially considering its potential downsides, its application in nanobot technology, its use in regrowing extinct plants, the potential for new diseases arising from its use, and the systematics of finding new plant species in places like the Amazon. Let's also consider the case of the dodo and the role of technology in its extinction. With a small sample size, AI can predict certain outcomes, a feature that can be beneficial in various fields. Most plant species are discovered rather than created, and AI can potentially help in predicting where these new species might be found. The question arises: is AI better at predicting the future? It can certainly help us see larger scale patterns that we aren't aware of. However, the act of predicting the future can create its own issues, akin to the Oracle of Delphi dilemma. For instance, can AI predict the third-order effects of its own intervention? By revealing patterns, AI becomes a more effective tool. The more layers of patterns it can show us, the better. AI and Medical Imaging: AI is increasingly being used in medical imaging, particularly through deep learning techniques. These have applications in MRI, CT, and PET scans, enhancing image reconstruction, quality, and efficiency. While impressive progress has been made, the technology still needs further development before it can be widely applied in clinical settings Transformers: Introduced in 2017, transformers are a type of deep learning model used primarily in natural language processing and computer vision. They're distinguished by their use of self-attention, enabling them to process the entire input data all at once, rather than sequentially as in Recurrent Neural Networks (RNNs). This allows for more parallelization and thus reduces training times. Transformers have become the model of choice for many NLP problems, replacing RNN models such as long short-term memory (LSTM)

In this podcast episode, the host, Stewart, is talking with Zach and Matthew, co-founders of Clay, a relationship management platform. The conversation covers various topics including: The Journey to Clay: Zach and Matthew discuss their journey of conceptualizing and developing Clay. They reveal the thought process that went into creating a product that could help manage relationships without reducing them to transactional entities. Exploration of Clay: They delve into the workings of Clay, its AI assistant, and how it helps users manage their relationships. Clay has the capability to remind users of their connections, keep them updated with their social interactions, and help maintain the quality of their relationships. Concept of Attention Economy: They talk about the notion of the attention economy and how it can be detrimental to our relationships. They also discuss how Clay can help users regain control over their attention. Relationship with Nature: Stewart discusses his personal experience with his move from San Francisco to the country, and the challenges and learning experiences that came with it. He prompts the others to share their thoughts on the relationship with nature, and how it influences our relationship with ourselves. Matthew relates this to the idea of self-discovery and the inherent need for human connection with nature. Adapting to Challenges: They discuss the challenges and "curveballs" of life, especially in the startup world, and how they adapt to them. Matthew highlights the importance of having a supportive co-founder and a solid team to weather the inevitable storms. Future Plans for Clay: Zach and Matthew hint at upcoming features in Clay, including AI advancements. They invite listeners to check out the beta preview of these features on their website. At the end of the conversation, Matthew emphasizes the importance of investing time in our relationships, not only in a professional context but also personally. He suggests that this investment offers a high return and could significantly improve one's quality of life.

Bart is the founding Guru at GuruScan | International Knowledge Management speaker | Makes Knowledge driven business decisions and helps enable the customers to do https://www.linkedin.com/in/bart-guruscan/ Guruscan website https://guruscan.nl/ What is knowledge management? Knowledge is a lot more than information Information is content Knowledge is explicit and implicit knowledge Knowledge Management strategy Shell Connecting people to people, Connecting people to content Community of practice Lessons learned, what did we do and how well did we do it? Forward-looking thing: how can we integrate learning and development? Learn about where we want to go This makes me excited to be a part of KM Skills that are really hard to make explicit Fingerspettein Riding a bike, talk to someone about riding a bike Keep on pedaling, look forward, and find your balance, these are all processes that can't be taught You can’t read how to ride a bike What percentage of knowledge for business is like riding a bike? Specific use cases; a lot of research 20% is explicit and 80% is actually stuff people are doing Then not ending up in the final 95% implicit in the particular case of tacit knowledge Is the role Thousands of people; how do you communicate with them Complex environment and things are changing in Solving complex problems is when you want to get people together Prehistoric groups There was cross-group and collaboration Strangers interacting 1-2 years now for 10-15 years of experience as specialist After a while, its interesting to hear how people have feelings about whether things are wrong Intuition says something is wrong, and then finds the thing that is wrong No textbook is going to tell you what is wrong Concept is called Dunbar number, robert dunbar, British anthropologist 150 people; the people with you can have a meaningful relationship High school friends are replaced with work friends Changes over time but the limit Social grooming, what their parents are doing, what are they doing If you want to expand you are not going to be FDR had 44,000 people in The level of leadership changes, and remote work As a CEO of a 20K person company Methods for Organizational network analysis Knowledge Map of the organization Connect people with very similar of knowledge. Find people to really like to exchange with An idea is network Bart is in Amsterdam Not totally remote Gitlab as an example Remote work as asynchronous Being able to work asynchronously in productive Large organizations Monday morning you have the standup Large organizations in tons of synchronous meetings Lockdowns the whole workspace Feeling productive vs not feeling productive Status report Alignment and updating people That's the big challenge The furthest in Async first Async needs to be changed If you can’t have that meeting, what would you do? Internal organization A lot of people who make money running the organization IIf you are up to 60 or 70 people because there is no overhead If you need to arrange something you need to Staff departments at 150 Institutional Staff departments Especially, growing the company as an incentive How do you work smarter, not harder? Our department In organizations, the hard thing is to make sure that you don’t reduce complexity, If you reduce the complexity Requisite variety, adapt to all the changes that are coming from the outside of the world Balance exploration and exploitation If you don’t exploit then you don’t the money Exploration is the future of the company How much money, time, effort and people? How much money should we invest in R&D? Insane amounts of money Every company should do in more exploration Changing processes is usually not considered R&D Changing your organization to better fit future Political aspect Produce 50 or 60% of all the semiconductors Flat screens LED lights have semiconductors European Union has different regulations Huge fabrication tension of where ASML TSMC The flow of money spent on the Governed by Moore laws The number of transistors on a square meter would double every year Fit the developments into the computer chips Pentium processors went faster than Moore’s laws How many people work in semiconductors? Ultraviolet lights Collaborations Semiconductor stuff, how to do the knowledge management? Work together with SAIS, German lens company SAIS maybe made an investment in that Seimer integrating with the equiptment Global recruiment that they do Optical engineering Thats the most important thing With customers and suppliers Crash in 2009 and 2020 Apple, Intel, and Samsung Flagship model The chain is so fragile

Matt Bunday works in crypto. He loves to rock climb, martial arts, and think about underground psychedelic therapy   What is the biggest problem you’ve faced with knowledge management at your companies? Slack  Dropped balls in communication LLMs might be the response Why do you think slack spread so fast even though its not the best product? It was a step up from email IRC was a component More friendly for non-engineering What do you think about the complexity of slack? Twist is an alternative Why is information architecture such a challenge? How do LLMs fit into this? What would happen if slack created a LLM or plugged one in? Slack workspace plug in What is a retrieval plug in? Universal adapter for any type of data Who are the incumbents in the slack knowledge management space? Guru startup What is the difference between information management and knowledge management? Knowledge management is a higher level synthesis Information management is siloing related types of information Data types  Group related to information around people What are your thoughts on the membrane? Siloes Privacy is where LLMs can be very innovative If we were to share a LLM We can both specify a privacy policy to the LLM and it will follow it LLMs can intuit the privacy public distinction Are you using LLMs to code for you? Copilot Issues with difficulty to prompt it correctly You had to write comments to prompt it Inline suggestions were not good Is it better now? Haven’t noticed a dramatic improvement Hard to prompt it to code in certain styles now GPT 4 is way better for a starting off point for projects Helpful for conversion processes What are the things that GPT4 has not been helpful for you? You have to chunk it What about building systems with GPT4? Code completion cool called tap9 Train the model against your local code What are some other things about KM that we can use tools for? Shared LLM for the family Surface serendipity between users If facebook were to do this One person says they are selling a couch One person is buying one LLM connects them At what point do we merge with the machines? Sufficiently high bandwidth Translation Are we already cyborgs? It began with wearing shoes Horselike Where do you fit in the excitement vs fear when it comes to AI? What parts of knowledge work will get automated? What are we losing? What is your take on bodywork? 10% investment Martial arts practices, he gets beat up  Ninjiutsu Special forces for ninjas Healing and striking points on the body are the same Balance between healing/killing, if you know how to heal Unbroken transmission since 1400 What is the importance 

  Zohar Atkins He is a rabbi, philosopher, blogger and podcaster. His podcast is called Meditations With Zohar Show Notes Who is your favorite philosopher? What is non-aggressive socratic questioning? Why do you like Heidegger? How was he Nazi-ism How do you relate to the history of philosophy? How do you walk the line between tradition and being a revolutionary? What was your first experience learning for yourself instead of learning top-down? What does it feel like to have a hunger for learning? What happens when we decide we are going to learn everything and then run into the block of only having so much time? How do you relate to patience when confronting the weird language of philosophy? How do you define good communication? What is the value of incomprehensibility? How important is banging your head on the wall? Who is a philosopher that you think is a fraud? Was Jesus a philosopher? Do philosophers build truth structures? What is hagiography? What is your philosophy of technology? How is Socrates exceptional? Where do philosophy and religion meet? What happened to the public intellectual? Was Wittgenstein religious?  How was Wittgenstein obsessed with language? What is your take on rationalism? What about scientism? What is the job of philosophy? Why are most people not interested in philosophy? Who is Leo Strauss? Philosophy is opposition to the state What happens when the state get too powerful and the philosophy gets crowded out? What is the dominant philosophy of the US in the 20th century? Pragmatism What was the difference between philosophy and science for ancient philosophers? How is philosophy a technology? How is UX a strain of phenomenology? What is the feedback cycle between technology and philosophy? What is the problem of induction? Aristotlean is ok with doing case studies Deductionism leads to cancelling all the case studies What defines the modern essence of technology? What happens when humans commoditize themselves? Without technology why can’t most understand leisure? How does Science doesn’t think? How did the original science people become more humble about the origins of science? How does achievement distract from the question of meaning? Why don’t you think Scientism is a big deal? How is Scientism is bad for religion? What is your take on new age occult stuff? Irrationalism sees with a squint to what rationalism is blind How can we become open to the strangeness of the universe? How can we be epistemologically humble?  

AJ is a software engineer interested in Urbit and data composability.  Twitter: https://twitter.com/ajlamarc https://www.ajlamarc.com/ Keep an eye out on Holium, and making it easier to code in Javascript for Urbit Notes What is data composability? Multiple applications sharing data and make the interfaces work together OpenAIs that both Twitter and Facebook had and then were shut down Data is the main product that these applications have You can separate the data and the code Personal AI How did we get to such a fragmented application landscape? Legacy internet vs New Internet? What is AJ working on? TomeDB Javascript package Associated gall agent NoSQL database The front end side is the interesting part What are gall agents? Standardized backend of Urbit Rules for moving data between different servers and accounts What are you looking for in a backend? Why are you reinventing the wheel when building a new urbit backend? What is the promise of urbit? What is the main draw of urbit? What does it mean to be permissionless? Urbit brings the best out of the past How much data is stored when you host your urbit on the cloud vs your own database? What are the technical limitations that an Uribit can store? 4-8 GBs can be stored and its stored in RAM How long until we have video streaming on Urbit? What is your take on where AI and Urbit mingle? How do you have an AI work for you rather than a big corporation? Urbit is creating a virtual world where you get only what you want and little of what you want The success of Urbit is the apple OS with the next generation of software How do you get the cost down? How do you run an urbit inside an urbit? What is the main difficulty of scale for computing on the cloud? If a million people join urbit tomorrow would it break? What is the biggest scale that Urbit has seen? 1-2K people What is the new narrative of Urbit? How do you use urbit as a really simple use case that urbit can solve to start getting urbit adoption? How do you store data and transfer it between two people in a secure way? Urbit as an infrastructure that can help build infrastructure There has to be the normal infrastructure that we are used to in terms of databases and networking  Urbit is mostly a backend technology Urbit could eventually have HOON native AI Did you use crypto to pay for anything in El Salvador? What did you learn at the Volcano Summit in El Savlador? Logan’s talk on Zorp using Knock  

Dave Snowden is the Chief Scientific Officer of Cynefin. Head of Knowledge Management for 30 years. Author of Cynefin - Weaving Sense-Making into the Fabric of Our World (https://tinyurl.com/4958x362)   When did knowledge management start in the 90s? The ultimate disciplinary field What is intellectual capital? Intellectual property What is the anarcha book? Knowledge management is information management What is relevant knowledge? What is messy coherence? What is exaptive innovation? How do you add value to organizations? Focus What is the difference between teleological idealism and realism? What is the KM process? Find out what is keeping middle management awake at night Do not want to be a CEO pet project How do you map what a company already knows? How do you map current knowledge from the things that keep people awake at night? Where are we on the cycle now? At the early stage of the hype cycle What is complexity theory? How do you map the unknown unknowns of a company? How do you create resiliency within an organization? How do you build informal networks across the organizations? European field guide on complexity management Does getting involved with tactics take away from strategy? Trying to make the cost of virtue greater than the cost of sin? How has knowledge management changed with Covid and remote work? How do you replicate pheromones in a remote environment? What are hexis? How does knowledge transfer work? How do you make decisions that keep options open? How do you create processes that stop ambiguity? Why are stories of linear processes greatly exaggerated? How do you deal with too much information? Focus on connecting people and storing information Entangled trios with task from different groups Run that every three months Secret is not to take an information centric approached Knowledge is only ever volunteered, not conscripted We always know more than we can say and we can always say more than we can write down What is necessary ambiguity? What is the role of narrative when it connects tacit to explicit information? Narrative asks you questions that make you think differently Lessons learning rather than lessons learned 90% of knowledge is walking out the door The danger of machine learning is dumbing down how we know things THE RIGHT SOURCE DATA IS THE KEY Machine learning is inductive  Feed ML better training data How did you get the role at IBM at knowledge management? Institute for knowledge management at IBM IBM center for complexity studies  How do you measure knowledge management? outcome/ouput measures Fine with predictable systems Outcomes produce a perverse incentive Vector measure for intensity of effort How do you structure KPIs? What are the power dynamics of exchange? Fear of abuse is the main reason people seek knowledge in organizations Art comes before words before human language Semiotics is symbols and signs UK is the most mapped country in the world We need renaissance instead of an enlightenment What is catholic with a small c?

Adam is the  VP of engineering at Invisible Technologies and active angel investment If you have specific questions, send them on to Adam@invisible.co   How do you organize information and data in order to spend most of your time on high-level work? How do you manage people who report to you to also do the same? What does knowledge management mean to you? What are the applications that will leverage LLMs? What is an intranet? Transition from card catalog to Google How do you take structured things and make them unstructured? Can LLMs do unstructured data? What is tabular data? Who is going to shine when it comes to leverage LLMs? How do LLMs hallucinate? How can we prevent LLMs from hallucinating? What would happen to a law firm that has an LLM that hallucinates a contract? What happens when NYC opened up all its APIs? What is data availability? Will LLMs have big moats? (13 minutes in might be when we get too close to OpenAI) What will happen with LLama from Meta? What has been overhyped in terms of LLMs? Have you started to use Copilot in your own programming? Reference tool rather than programming aid LLM code has a higher security risk What is your take on AI ethics? How do you deal with collective commons type of stuff? How does bias training come into play? How do we think about the implicit biases that the models have been trained on? How does technology help us to understand ourselves? How is technology an amplification of human ability? What is the frustration of knowledge management for you? Bot at Facebook that goes through and warns people when something needs to be updated Finding something and its wrong How do we make sure that information stays updated and relevant What is the curse of knowledge? Onboarding is testing the system How should we think about knowledge leakage? Does Notion have an API? Keep metrics on how much people are writing What were some ways that Facebook excelled at knowledge management? Investment in search tools Code documentation Chat bot messages scale of 1-10 how do you feel   What are the barriers that we have at invisible to implementing great search? What is the issue with implicit knowledge at Invisible? What is status hero? 8 or 9 engineering teams at invisible When there is high trust all processes work and when there is low trust no processes work What is the biggest problem going to be for me as knowledge management guy? How can we bring spaced repetition software into Invisible?