Fernando from Trend Micro built AI voice agents that simulate customer calls to train their sales teams. We dive into how he went from a language learning chatbot to enterprise sales training with voice AI that gives real-time feedback and scores your performance.
A conversation about how LLMs are automating our ability to patch vulnerabilities and a look into the future in a world where software can heal itself.
Video walk-through: https://youtu.be/6VUI1buwbdY
Blog: https://www.dylandavis.net/2024/11/self-healing-code/
Blog post: https://www.dylandavis.net/2024/07/three-gpts-walk-into-a-bar-and-write-an-exec-summary/
TL;DR: This blog covers three custom GPTs I created to help with creating executive summaries: one extracts insights, another drafts the summary, and the third refines it for clarity. But! I still recommend using writing as a tool for critical thinking.
Blog post: https://www.dylandavis.net/2024/05/simplifying-outlive-to-extend-your-healthspan/
TL;DR: This blog post breaks down Outlive by Peter Attia into actionable steps for those wanting to improve their healthspan. It covers Attia's "Medicine 3.0" approach with practical tips on exercise, nutrition, and sleep to help readers stay active and energized well into their later years.
Blog - https://www.dylandavis.net/2024/01/life-lessons-i-wish-i-knew-at-17-a-letter-to-my-niece/
TL;DR: An uncle offers candid life advice to his teenage niece, emphasizing the importance of practical career choices, lifelong learning, and critical thinking. He highlights the value of self-discipline, networking, financial literacy, and recognizing life's reversible and irreversible decisions. The message is to be ambitious yet skeptical of advice, including his own, and to understand that adults are also continuously learning and adapting.
Blog post: https://www.dylandavis.net/2024/01/openais-2024-decline-a-value-prediction/
TLDR: This post discusses Chamath's prediction of a 2024 dip in OpenAI's value, focusing on industry trends like improved AI model quality, faster response times, and lower energy costs. It highlights the rise of efficient computing chips and the split between large foundational AI models and smaller, more efficient ones
Blog post - https://www.dylandavis.net/2024/01/how-do-you-keep-up-with-ai/
In this blog, I share my top AI resources, categorized and ranked to stay updated in the field. From YouTube channels to podcasts, each is chosen for its unique insight and value in AI. This guide is designed to help enthusiasts easily navigate the world of AI.
Blog post - https://www.dylandavis.net/2023/12/disappearing-data-moats/
This post explores the shift from traditional data moats to synthetic data in AI. It discusses how synthetic data, generated through methods like GANs and VAEs, is challenging the reliance on consumer data by offering diverse, high-quality alternatives without privacy concerns. The piece highlights synthetic data's growing importance in AI training and its potential to drive future advancements in the field.
Blog post - https://www.dylandavis.net/2023/12/securing-the-new-era-of-ai-driven-operating-systems-a-novices-tale/
The blog post explores the concept of Generative AI (GenAI) as the foundation of future operating systems, replacing traditional components with AI-driven counterparts like 'Orchestrator LLM' and 'Context Window'. It discusses the shift in user interaction towards voice and gestures, highlighting advancements in AI technologies like ChatGPT and Apple's Vision Pro. The post also addresses security concerns in GenAI OS, suggesting adaptations of traditional security measures and introducing new strategies tailored for AI-driven systems.
Blog post audio - https://www.dylandavis.net/2023/12/genais-shift-from-cyber-villain-to-llm-protector/
Unraveling the brighter side of Generative AI (GenAI) in cybersecurity, a realm often overshadowed by its darker uses. In this episode, we're flipping the script from doom and gloom to showcasing GenAI as the cybersecurity defender's new best friend. We're diving into the latest GenAI tools tailored for safeguarding LLMs from cunning cyber threats. We'll walk you through the nuts and bolts of LLM firewalls, the avant-garde concept of control tokens, and breakthrough developments like Meta's Purple Llama.