High Bit

“Robots are the only way my business survives, but it’s not viable for me.”Formic founder and CEO Saman Farid joins Brett Gibson, managing partner at Initialized to unpack why that mindset keeps factories from adopting automation and how Formic closes the gap. They cover: de-risking with financing, productizing complete robot work cells, and running fleets with teleoperation, intelligent error recovery, and careful staging to hit factory-grade uptime. You’ll hear why palletizing is the ideal first beachhead, how the team cut deployment costs roughly in half, and why they say no to one-off requests until they can be productized. Saman also shares how the company resists “fun” engineering in favor of scale, injects controlled chaos into his company, uses daily 8 a.m. meetings for problem solving, and bridges the culture gap between the manufacturing and software industries.Subscribe for more builder-level deep dives from High Bit.Follow Formic and Saman for more:Formic: https://x.com/goformic Saman: https://x.com/samanfaridContent(00:00) “It’s not viable for me”—closing the adoption gap(00:44) What Formic does & where robots work today(02:24) Engineer → VC → founder: why start Formic(04:38) Adoption vs flashy demos. Solve one task well(05:47) De-risking with financing; manual first, then automate(08:23) The playbook: scope, build modules, deploy, operate1(10:36) Work cells, not just arms (13:38) 99.9% uptime15:01 Why palletizing was the first beachhead(16:49) Cutting costs per deployment(18:08) Saying “no” & expanding scope the right way(21:15) Resisting “fun” engineering to serve more factories(22:07) Injecting chaos into your company(23:34) Daily 8am to crack hard problems(25:40) Culture clash: manufacturing × software(27:24) Evaluating new robots, regional rollouts(31:24) Where AI helps across the org(37:11) What’s next: more robots, more tasks, more factories

Brett Gibson hosts Sol’s Ben Chelf and John Boiles on the "High Bit" podcast to discuss their groundbreaking work on the Sol Reader, the world’s first wearable e-reader. They dive into the unique engineering challenges of integrating e-ink technology into a head-mounted format, the intricate development process of creating their own e-ink driver, and the importance of optimizing for user comfort and the reading experience.Chapters:00:00 - Introduction02:16 - The Birth of Sol Reader: From Idea to Prototype05:19 - Comparing Sol Reader to Other Wearable Technologies09:59 - Technical Challenges and Innovations in E-Ink Displays20:30 - Manufacturing and Quality Control Processes23:12 - Software Development and Optimization33:40 - Future Plans

Vince Chu, cofounder of HomeVision, joins the "High Bit" podcast to discuss how his team is automating real estate valuations at scale. From overcoming technical challenges to scaling AI-powered underwriting, Vince explains how HomeVision is changing the game for both lenders and appraisers.

Our managing partner, Brett Gibson, spoke with Playground AI founder Suhail Doshi about advancements in AI-driven image generation and their new Playground v3 image foundation model. Suhail shared Playground’s mission to make graphic design more accessible, enabling users to create items such as T-shirts, logos, social media posts, and even memes. They discussed the gap between language and image models, as well as the underinvestment in AI for images. Check out the new Playground here: https://playground.com/design Playground’s white paper introducing their image foundation model Playground v3 here: https://arxiv.org/abs/2409.10695

On the latest episode of our podcast “High Bit,” Brett Gibson spoke with Pratap Ranade, CEO of Arena, about the intricacies of using AI to convert data into real-time simulations for optimizing sales, pricing, and marketing strategies. Pratap explained how Arena navigates the complexities of large-scale enterprises, including those in the consumer goods and advanced manufacturing industries. He also shared insights from his background in quantum physics and machine learning and discussed future partnerships and product developments for Arena. 00:00 Introduction 00:47 Meet Pratap Ranade, CEO of Arena 01:58 Understanding Arena’s core functionality 02:44 Challenges in building Arena 05:03 Consumer goods and advanced manufacturing 07:09 The importance of abstraction and metadata 13:50 Pratap's journey and vision 19:35 The future of Arena

In this episode of "High Bit," Initialized partner and Rent the Runway co-founder Jenny Fleiss discusses how she developed a proprietary logistics system for a consumer brand. Jenny opens up about the early challenges faced in starting Rent the Runway, the strategic decisions made in scaling a complex operation, and how she built an in-house logistics team. CHAPTERS 00:00 Introduction 1:45 Jenny's transition to investing and current areas of investing interests at Initialized. 4:38 The initial logistics challenges at Rent the Runway. 6:36 Partnering with local dry cleaners and scaling operations. 9:01 Building a proprietary logistics system from scratch. 13:00 Key takeaways from managing and scaling a complex logistics operation.

In this episode of "High Bit," Automat founder and CTO Gautam Bose talked about the challenges of working with LLMs in robotic process automation, and the importance of rigorous testing procedures. CHAPTERS 00:00 Introduction 01:30 What Automat Does and How They Started 03:06 Challenges in Building RPA 08:46 What the Future of RPA Looks Like 12:52 How Automat Ensures Reliability in Its LLMs 14:53 Automating Complex Workflows With Robust SOP Collections 20:17 Coming Up With Solutions to Complex Orders 23:41 Improving Ideas Through Iteration

In this episode of "High Bit," Peter Kieltyka discusses the dual challenge of maintaining decentralization and accessibility in blockchain technology, and the growing significance of APIs in building connected systems. CHAPTERS 00:00 Introduction 01:38 What Horizon Does and the Challenges in Delivering the Crypto Gaming Platform 04:53 How Peter Came to the Decision of Working on Blockchains 07:07 How APIs Are Created for Modern Client Server Web Applications 12:25 All About GRPC 15:40 Chi: Peter’s Software That Lets You Write Microservices 21:15 Peter’s 2 Favorite Books He Recommends to Everyone He Works With 27:27 What Else to Look Forward To from Horizon

In this episode of "High Bit," Brett Gibson chats with Kurt Mackey, CEO of Fly.io, to break down how they challenge big cloud providers, overcome global logistical nightmares, and continuously evolve their software stack. From hardware setup challenges to growth-driven solutions, Kurt explains the highs and lows of managing a developer-focused public cloud. 00:00 Introduction 1:21 Starting off with what Fly.io is 3:26 Zooming in on the developer experience 6:19 What Fly.io’s journey was like 19:23 Key takeaways from setting up the public cloud globally 25:20 What we can expect from the developer-facing stack trend

In this episode of "High Bit," host Brett Gibson interviews Rahul Raina, cofounder and CTO of TRM Labs. Rahul shares insights into TRM's mission of creating a safer financial system and discusses the launch of their innovative features. The conversation delves into Rahul's background, TRM's approach to tackling financial crimes, and the challenges of processing massive amounts of data in the ever-evolving world of crypto.  00:00 Introduction 2:09 TRM’s mission to build a safer financial system 9:20 The question that started TRM 14:47 Breaking down their iteration process 20:21 The journey to pushing the state of the art 24:52 Proportionality as update strategy

In this episode of "High Bit," AJ Asver, CEO of Parcha, discusses the challenges of building AI agents for complex tasks. Parcha's AI agents automate operations and compliance, addressing the gap between proof of concept and production. 00:00 Introduction 01:51 Introducing Parcha and what they do 4:02 Breaking down how an AI agent relates to LLM 12:53 The challenges in scaling AI agents 17:15 Making the system as accurate as possible 28:51 Predicting the rise of verticalized agents

In this episode of High Bit, Aaron discusses the challenges of hiring a team to build an infrastructure-as-a-service platform for blockchain networks. When Aaron and his cofounder Joe Lallouz first created Bison Trails, they initially faced challenges running nodes and infrastructure for different networks. They found that by investing in tests early on and improving automation capabilities, they could streamline operations. Coinbase eventually acquired Bison Trails in January 2021. For more info on High Bit, and to listen to all of the other episodes, visit: www.Initialized.com/HighBit Topics: 2:28 Aaron's Background and How He Became an Engineer 5:21 Building Grand St. and Selling the Company to Etsy 10:55 The Early Days of Building Bison Trails 16:21 Building the Early Team of Infrastructure Engineers with Blockchain Experience 19:53 Slashing: The Penalties Imposed on Validators in Blockchain Networks for Misbehavior 27:07 What Bison Trails Taught Aaron About Building Large Organizations 33:28 Aaron on Lessons He Learned While Hiring the Bison Trails Team 38:00 Life After Bison Trails and Coinbase. What's Next for Aaron Henshaw? High Bit is a new podcast hosted by Initialized Capital managing partner Brett Gibson about the art of technical problem-solving. A high bit is the most significant part of the binary representation of a number. In programming language, it is commonly referred to as the most important thing you need to understand about a problem. We spoke with our first four guests about just that. In each episode, we’ll break down a gnarly engineering problem and hear how the builder’s ingenuity and inventiveness led to a successful outcome. High Bit is produced by Initialized Capital and hosted by Brett Gibson. Our showrunner is Candy Cheng. Episodes were shot and edited by Jordan Berns. Production support by Kevin Heinz and BTB Video Production.

In this episode of High Bit, Jose walks us through the challenges of troubleshooting a complex problem encountered during the pre-launch testing of a space-bound asteroid mining refinery. Using a real-life example, he illustrates the application of a fishbone diagram and the necessity of methodical problem-solving, even under extreme time pressure. Topics: 1:41 What is AstroForge? 7:13 How Jose Met His Cofounder Matt Gialich 13:18 Using a Fishbone Diagram to Find the Root Cause of a Problem 15:01 The "Holy Shit" Moment When a Launch Does Not Go According to Plan 20:37 Validating All the Points of Failure to Get to a Solution 26:25 Lessons Learned from Working Under Time Pressure High Bit is a new podcast hosted by Initialized Capital managing partner Brett Gibson about the art of technical problem-solving. A high bit is the most significant part of the binary representation of a number. In programming language, it is commonly referred to as the most important thing you need to understand about a problem. We spoke with our first four guests about just that. In each episode, we’ll break down a gnarly engineering problem and hear how the builder’s ingenuity and inventiveness led to a successful outcome. High Bit is produced by Initialized Capital and hosted by Brett Gibson. Our showrunner is Candy Cheng. Episodes were shot and edited by Jordan Berns. Production support by Kevin Heinz and BTB Video Production.

In this episode of High Bit, Ethan reveals the unique challenges of building a reliable and global crypto trading platform. “There’s no downtime because crypto trades 24/7,” he said. “So in that way, it’s a lot more like Google than the New York Stock Exchanges.” Ethan also shared the art of balancing safety and speed in product development. “You need to adapt, and you need to change up your workflow as the market changes. We want to wrap that whole messy ecosystem in a layer that allows institutions to interact with crypto just like they would any other asset class.” Topics: 02:26 From Broadway Technology to Talos 06:23 Crypto Trading is "More Like Google Than NYSE" 11:46 Examining the Various Failure Scenarios of Trading Crypto 15:56 The Importance of Getting Details Correct in Making a Successful Trading System Work 21:35 Client Communication as Part of the Engineering Effort 25:05 Building Safety Features into the Product 29:30 Adaptability Around Messiness High Bit is a new podcast hosted by Initialized Capital managing partner Brett Gibson about the art of technical problem-solving. A high bit is the most significant part of the binary representation of a number. In programming language, it is commonly referred to as the most important thing you need to understand about a problem. We spoke with our first four guests about just that. In each episode, we’ll break down a gnarly engineering problem and hear how the builder’s ingenuity and inventiveness led to a successful outcome. High Bit is produced by Initialized Capital and hosted by Brett Gibson. Our showrunner is Candy Cheng. Episodes were shot and edited by Jordan Berns. Production support by Kevin Heinz and BTB Video Production.

In this episode of High Bit, Renuka shares the evolution of Clockwork’s AI-powered nail-painting robot. She shares her insights into safe AI utilization and how she adopted software engineering principles into the hardware process. Topics: 1:56: Renuka’s Background as a Systems Engineer at Nvidia 5:27: Using an Off-The-Shelf AI Model to Create the First Version of Clockwork 10:06: Discovering Pointillism in Nail Painting 14:46: Using AI in Safe Ways 18:29: A Very Humbling Lesson for a Software Engineer 22:55: Making Nail Painting as Easy as Grabbing a Cup of Coffee High Bit is a new podcast hosted by Initialized Capital managing partner Brett Gibson about the art of technical problem-solving. A high bit is the most significant part of the binary representation of a number. In programming language, it is commonly referred to as the most important thing you need to understand about a problem. We spoke with our first four guests about just that. In each episode, we’ll break down a gnarly engineering problem and hear how the builder’s ingenuity and inventiveness led to a successful outcome. High Bit is produced by Initialized Capital and hosted by Brett Gibson. Our showrunner is Candy Cheng. Episodes were shot and edited by Jordan Berns. Production support by Kevin Heinz and BTB Video Production.

Join Initialized’s Brett Gibson as he discusses thorny engineering problems with his guests on “High Bit,” a podcast about the art of technical problem-solving.

AI models keep getting better, but most AI systems still fail in production. Why?In this episode of High Bit, Brett Gibson sits down with Ghita Houir Alami, cofounder and CEO of ZeroEntropy, to break down the real bottleneck holding AI agents back: retrieval.Ghita explains why embeddings alone can’t reliably surface the right information, why tools like Slack search feel so frustrating, and how rerankers add a critical second pass that dramatically improves accuracy. She walks through ZeroEntropy’s approach to training rerankers using pairwise comparisons and Elo-style scoring, and why this method generalizes across domains like code, finance, and biology.The conversation goes deep into:Why AI agents fail even when the data exists.How reranking fixes poor ordering from vector search.Why “accuracy” now includes helpful context, not just correct answers.What actually changes when retrieval becomes trustworthy enough to remove humans from the loop.If you’re building AI agents, search systems, customer support bots, or internal knowledge tools, this episode explains what’s breaking today, and what has to change for AI to work reliably at scale.(00:00) What changes when retrieval works(00:39) What ZeroEntropy builds(01:42) Why retrieval became the real problem(03:12) Why search fails (Slack included)(05:11) Why embeddings fall short(07:11) Rerankers: the missing layer(10:11) Why rerankers matter most(12:44) Pairwise ranking vs scoring(13:52) Elo scoring for documents(16:33) Fast rerankers via distillation(18:07) Why old training methods break(21:29) Retrieval for AI agents(24:20) Recency, memory, personalization(32:06) What reliable retrieval unlocks(33:42) What’s next for ZeroEntropyFollow Ghita and ZeroEntropy for more:X@ghita__ha@ZeroEntropy_AILinkedInhttps://www.linkedin.com/in/ghita-houir-alami/https://www.linkedin.com/company/zeroentropy-inc

Electrification isn’t easy—and most people don’t see the chaos beneath the solar panels, batteries, EV chargers, and heat pumps going onto the grid. Coperniq cofounder and CTO Max Kazakov breaks down the hidden workflows behind distributed energy: legacy tools installers still rely on, hardware that doesn’t want to integrate, and why the next generation of “utilities” will look nothing like the last.Coperniq is the workflow platform for contractors and energy companies to move from post-its and spreadsheets to a system that sells, permits, installs, and maintains distributed energy assets over decades.Max also shares what it takes to build vertical SaaS for the physical world: curbside Figma demos during COVID, rebuilding their mobile app for 120°F rooftops with no cell service, designing a workflow engine that matches real-world permitting and interconnection, integrating a wild west of OEM hardware, and how AI is already reshaping their product and engineering culture.Content:(00:00) The Invisible Glue of the New Grid(01:05) The Second Electrification Wave(02:51) Cofounder Origins: Russia, Yemen, Berkeley(06:12) Humans + Hardware Coordination Challenge(08:10) Anti-MVP: Mini ERP on Day One(11:56) Curbside Figma Demos during COVID(14:47) Field Reality: 120° Rooftops, Zero Cell Service(20:03) Stateful Workflows (Permits, Interconnection, Construction)(24:56) Integrating OEM Hardware (Hitting Walls)(28:41) The Dongle Question: Do we need software afterall?(31:25) Rebuilding the Mobile App for an Offline-First World(36:40) Hire Tinkerers, Not Pedigrees(43:00) How AI Is Reshaping CoperniqSubscribe to High Bit for more conversations with technical founders building the future.Follow @Coperniq_AI for more.

Robots built like humans.On High Bit, Dhanush Radhakrishnan, cofounder & CEO of Clone, explains how they’re letting biology set the blueprint for musculoskeletal, super-intelligent androids — synthetic humans straight out of sci-fi.Powered by artificial muscles instead of motors and attached to anatomically accurate skeletons, Clone is building robots designed for human-level motion, durability, and full-body control.Dhanush explains the early engineering choices that helped them move fast, their data strategy (motion capture, teleoperation, egocentric video), and his excitement about a future untethered biped.Hosted by Brett Gibson, managing partner at Initialized.Content:(00:00) From Fragile to Durable(00:37) Musculoskeletal Androids(02:03) Why Artificial Muscles(04:10) Starting Clone in Poland(06:29) Removing Early Sensors(10:14) The Durability Challenge(14:10) Anatomy as Blueprint(16:22) Building Custom Valves(18:17) Hand to Full Body(24:40) Prototyping with Pneumatics(29:26) Delaying Tactile Skin(32:39) Data: MOCAP + Teleop(45:20) Toward an Untethered BipedSubscribe to High Bit for more conversations with technical founders building the future.

In this episode, Brett Gibson talks with Lucas Young, cofounder and CEO of Deepnight about how they’re building AI-powered night vision that helps the military, law enforcement, and first responders see in near-total darkness. Deepnight combines AI with commodity digital sensors — the same kind used in smartphones — to replace expensive analog night-vision hardware that costs over $30,000 per unit and hasn’t kept pace with modern imaging technology.Lucas explains how night vision has worked since World War II, why analog image intensifiers hit a ceiling, how smartphone photography paved the way for this breakthrough, and what it takes to bring military-grade low-light imaging into the field.Chapters(00:00) Why Night Vision Is Still Mostly Analog(00:39) Deepnight’s Breakthrough: AI That Sees in the Dark(01:44) How Their AI Reconstructs *Real* Scenes(03:58) Lucas’s Path: Google Pixel → YC Founder(05:26) Why Modern Cameras Rely on Software(09:12) The Rise of AI-Enhanced Photography(11:45) The Insight: AI Could Beat $30K Night-Vision Goggles(13:03) How Traditional Night-Vision Tubes Work(14:10) Starting Deepnight Without Knowing If It Would Work(15:11) Early Prototypes: Offline → Real-Time Night Vision(16:15) The Physics Challenge: Seeing in Moonless Starlight(19:12) Running This on Smartphone-Class Chips(22:27) Building a Custom Neural Network for Night Vision(28:43) Can Cheap $50 Sensors Match Military Gear?(48:06) What’s Next: Real Soldiers Using AI Night VisionSubscribe to High Bit for more conversations with technical founders building what’s next, hosted by Brett Gibson of Initialized Capital.