Hallway Chat

Today, Nabeel and Fraser tackle the questions they've been asking internally. What level of improvement would it take to disrupt an incumbent product? They also explore what advantages second movers have in product markets, the emerging importance of reasoning models and computer use products, and what makes certain legacy markets ripe for AI reinvention. (00:00) - Intro (01:28) - Can 10% better, year after year, win? (10:43) - The questions we are asking (11:11) - Second Mover Advantage (26:55) - Exploring Deep Research and Competitors (29:11) - The Future of Computer Use (35:56) - AI as Muse: Enhancing Human Creativity

Nabeel and Fraser sit down for a round-table discussion with Chris Pedregal, founder and CEO of Granola, an AI note taking app, to discuss Granola’s evolution and product philosophy. Their conversation touches on the challenges of developing user-friendly AI products, and how to balance that with product vision. They also talk about what it’s like being a second-time founder, as well as if and when AI wearables will gain widespread adoption.  (00:00) - Christopher Pedregal - Granola & the Art of Invisible AI (01:05) - Second time founder syndrome (01:42) - The path to Granola (04:13) - Starting by "playing" (11:16) - The friend test (12:25) - Product philosophy behind Granola (16:25) - AI as Augmentor (22:04) - Balancing beginner and expert users (34:21) - Granola's Initial Launch and Challenges (34:52) - Balancing Product Development and Market Fit (35:58) - Where Granola is headed from here (40:00) - Vertical AI vs. General Assistants, who wins where? (42:44) - The Wisdom of Experts Era (50:54) - AI passing judgement (53:51) - Wearables role in AI (01:00:25) - Social Dynamics and Privacy Concerns (01:02:32) - The Need for New Norms and Labels (01:03:01) - What else could be Granola-ized

What lessons will we take from 2024's AI products into next year? How better to reflect on advancements in the field of AI over this past year than to ask AI itself? Today, Fraser and Nabeel use AI models like ChatGPT, Claude, and o1 to first look back and see if products and innovations of 2023 lived up to their hype in 2024. And to look at the show’s year in review, Nabeel puts these AI models to the test by feeding them episode transcripts and seeing which model produces a usable summary. Together, they also dive into the effectiveness of Google Gemini’s Deep Research tool, the rise of agentic computing in 2024, and whether products ever fundamentally change or just get constantly re-invented.  (00:00) - Introduction (00:48) - Reflecting on a Year of AI Releases (00:23) - The Rise of Agentic AI (00:38) - Google Gemini’s Deep Research (00:47) - Using AI Models to Review 2024 at Hallway Chat (00:56) - The Role of Taste in Making Great AI Products (00:06) - The Rollout of o1 (00:06) - WebSim - Self Expression Through Software (00:36) - Same Products, Different Iterations (00:39) - A Second Look at Apple Intelligence (00:10) - Living Agentically with AI (00:18) - Final Thoughts

Everywhere you turn, companies have shoved AI features into their products that no one is using. Today, Fraser and Nabeel ask each other—are there any AI features that you’re actually using in older products? Plus: How kids will be shaped using AI products daily. What would a fully unstructured Slack look like. AI email habits. Do we think incumbents will catch up with AI integrations? PREM framework for startup prioritization. Is this startups vs incumbents, or just startups vs the big llms? (00:00) - Intro (00:52) - Welcome to Hallway Chat (01:05) - What have incumbents really done with AI? (08:59) - Trends in AI integrations that work (20:21) - Reflecting, it's still early (23:42) - Creating AI that manages up (28:47) - PREM: A framework for prioritization (34:32) - Forget incumbents, how do startups fight fend off the LLMs? (37:24) - What are the new horizontals?

Open format Q&A this week. Fraser and Nabeel explore AI data privacy, the ethics of copying features, and maintaining innovation. They discuss enterprise data challenges, the importance of a strong product identity, and strategies for early-stage startups during fundraising season. (00:00) - Open (00:53) - Q&A Session Kickoff (01:14) - The data you gather is your roadmap (14:03) - The gravity of slowing down in your startup (21:20) - Your quarterly goal: Something Fundamentally Changes (21:57) - Low Risk, Low Outcomes (24:31) - Large Organizations and Mediocrity (30:58) - When to steal a feature?

Join Fraser, Nabeel, and special guest Anil from Meter as they dive into the innovation behind Meter Command. Meter's new interface has been called the "future of software" and a far ranging interview talks about the problem of designing product around personas, how command interfaces bridge the gap between CLI and dashboards, the importance of owning your tech stack, how your data is your product roadmap, and the orientation to long term thinking while still staying on the cutting edge. Links* See a demo of Meter Command* There's a lot of fine work from Bret Victor on visualization and interfaces, but here's one to get you started* Geoffrey Litt's Malleable SW in the age of LLMs* Hunter Walk's post on OKRs post his stint at Google (00:00) - Intro (00:19) - Meter Command and the future of software (03:36) - What we forget about software making (06:39) - Models make you reassess your own product (07:27) - Why aren't there more AI explorations in software? (09:38) - The "efficiency era" of startups (13:21) - Software should be soft (17:13) - The problem of "persona building" software (21:29) - Enterprise software is reporting (22:37) - Your data is your product roadmap (25:13) - Emersing in the trends of AI as a founder (27:27) - From janky prototype to custom model building (29:55) - "No AI" - Build for the problem, not the technology (36:50) - Full stack first, APIs over time (40:14) - If you want to be early you have to be full stack (48:28) - Long term orientation (50:56) - What one thing would you work on now (55:41) - End

How would you design an AI product differently for a hobbyist vs prosumer vs professional? Using MidJourney as a case study in the passion economy, they delve into how AI tools can help build communities and support semi-professional creators. They explore how these tools can democratize creativity, the role of art vs craft, the role of grit in skill building, cheating vs copiloting, the missing middle of the creative markets, and more. They also touch on the broader implications of AI in different creative fields from music, writing and podcasts. It's an exploration of how AI can help users maintain creative flow and the underserved opportunity for AI to help unleash the hobbyist. Nabeel also spoke on Midjourney's unique growth loops on Unsolicited FeedbackLe Jin has great writing on the passion economy and wrote about the creative middle classLinus has an excellent article on Epistemic calibration that touches on taste, that we didn't get in the episode but worth seeing. Simply Drawing - ai feedback meets artRobert Morris' the Power Broker, which touches briefly on the rise of the professionalized classa little more on why bees wiggle outside of near-term evolutionary needs00:00 Intro00:52 AI in Creative Endeavors01:58 Market Expansion03:01 The underexplored customer in AI08:23 Finding other AI hobbiest opportunities11:31 If they AI made it, are we building skill?13:36 Difference between art and craft14:40 Homework Helpers vs Writing Copilots16:07 Hobbiests are not quite the creator economy17:47 Product for hobbists21:03 The missing middle of creative markets26:42 AI's potential in the education of craft (00:00) - Intro (00:52) - AI in Creative Endeavors (02:00) - Market Expansion (03:02) - The underexplored customer in AI (08:25) - Finding other AI hobbiest opportunities (11:32) - If they AI made it, are we building skill? (13:38) - Difference between art and craft (14:42) - Homework Helpers vs Writing Copilots (16:09) - Hobbiests are not quite the creator economy (17:49) - Product for hobbists (21:04) - The missing middle of creative markets (26:44) - AI's potential in the education of craft

Fraser and Nabeel start by retcon'ing their earlier "maybe we should stop talking about models" and reflecting on the importance of adapting product strategy based how much disruption is happening in your market. The episode highlights the release of Claude Artifacts by Anthropic, and comparing it to other AI product innovations. They also nerd out on the Xbloom coffee machine and a robotic mop. Lastly, Fraser asks about how to navigate when a founder gets introduced to the wrong partner at a VC firm. (00:00) - Product development early on the S curve, exploring Anthropic Artifacts (00:43) - Welcome to Hallway Chat (00:58) - Strong Convictions Loosely Held (03:37) - Exploring Claude Artifacts (10:48) - S Curve Product Design (23:33) - The Instinct to Exhale: Post-Launch Reflections (25:25) - Balancing Innovation and Stability (27:00) - Leaning into risk (32:30) - Question: Getting introd to the wrong partner at a VC

Fraser and Nabeel dive into the evolving landscape of AI note-taking apps, focusing on a new product, Granola. They discuss Granola's unique approach of "enhancing your thought" instead of "thinking for you." They also explore: applying this idea to other categories, the influence of LLMs on UX, and the need for more experimental user workflows to unlock AI. They also discuss the challenges and dynamics of startup funding, the significance of capital in achieving market success, and how dynamics of AI team building are changing. (00:00) - Introduction (01:05) - Introducing Granola: The AI Note-Taking App (02:19) - Granola's Unique Features and User Experience (04:54) - Challenges and Innovations in AI Note-Taking (06:53) - Granola's Practical Use Cases and Feedback (09:54) - Reducing Friction in Product Experience (13:04) - Where is the AI innovation? (16:11) - The Role of Incentives in Innovation (21:51) - When competitors raise 10x more than you (23:11) - Introducing Granola: The AI Note-Taking App (23:18) - Granola's Unique Features and User Experience (24:35) - Challenges and Innovations in AI Note-Taking (25:07) - Granola's Practical Use Cases and Feedback (25:19) - Reducing Friction in Product Experience (26:44) - Where is the AI innovation? (27:12) - The Role of Incentives in Innovation (27:59) - When competitors raise 10x more than you (28:34) - Overcapitalization and Its Consequences (29:21) - Customer-Centric Approach in Startups (30:22) - How AI Startups are Building Teams Differently (42:41) - Final Thoughts and Product Recommendations

Is it still worth it to discuss models when discussing startups? Nabeel and Fraser discuss how that may be the wrong question to ask in the current landscape, and why customer-centric questions and user experience should be the basis of product experience. Later, they deliberate who might come out on top in the “horse race” for AI product dominance, and whether it will come from a large, established company, or if the frontier of capabilities belongs to small innovators. (00:00) - Decoding the Future: Puzzles vs. Mysteries in Tech (01:22) - Welcome to Hallway Chat: Podcast or Tweets? (01:35) - Exploring the Venture Firm USV's "Hallway Chat" Tweets (02:27) - The atomization of media (03:17) - Rethinking the Focus on AI Models in Startups (04:38) - The Importance of Use Cases Over Models in AI Innovation (07:45) - What makes a Foundational Model... Foundational (18:23) - AI for Consumers: Navigating the S Curve of Innovation

What is the extent of autonomous coding engineer Devin’s ability to generate real, functional applications with little to no help? Nabeel and Fraser dive into the buzz about Cognition’s ‘Devin’, what makes it different, and the transformative potential of AI in software engineering, particularly focusing on autonomous coding software. Later, they get into the innovator's dilemma and the lessons the .com era can lend to this new time in AI.(00:00) Intro(01:23) The buzz about Cognition’s Devin(07:44) What makes Devin different?(12:19) Tolerance for time(16:07) The interface of the future(22:19) Innovating around the incumbent’s advantage(25:30) Cutting edge products mean new user bases(29:52) Netscape was the Open AI of the Mobile Revolution(33:42) Optimism as the engine of capitalism(37:56) The model is not the product

This week is a discussion of the points of differentiation vs commodification in various AI models. Including how these points might change over time, and might change between language, image, and video models. Then Fraser pivots to orienting around jobs to be done in AI, and how the various models have a huge gap between being capable of doing something, and doing it well. We then talk about Claude 3, the nature of benchmarking, and the rapid dropping LLM prices. Lastly, we cover a startup subject, debating the merits of SAFE (Simple Agreement for Future Equity) vs. priced equity rounds in early-stage funding. Links:Claude 3 from AnthropicLMSys Chatbot Leaderboard for which models are sticking out right nowFor areas where models don't differentiate, we are back to the 7 powers framework. Hamilton Helmer's website with overview of the 7 Powers framework: https://7powers.com/, Sachin Rekhi has a good detailed primer on the 7 Powers as well.Carta guide on priced rounds vs SAFEs, which we don't really agree with all the pros and cons, but is a decent overview (00:00) - Opening (00:31) - Are models an inevitable commodity? (05:58) - How image, video, and other models may pan out differently than language (08:20) - It's not the model, it's the customer (10:20) - Jobs to be done in AI - 1. Can it do it... 2. very well.. 3. exactly how I want it to. (22:13) - Claude 3 by Anthropic (25:25) - ELO Leaderboard Results (26:25) - Claude 3 Haiku and Falling LLM Prices (30:33) - SAFES vs priced equity rounds

Josh Miller, Founder and CEO of The Browser Co, joins this week. We loved the experience with Arc Search, a new AI-enabled mobile search experience, we decided to talk to Josh directly about how it all came together. Topics:* How Arc Search came to be spawned out of basically a side project* The core value of "we don't know"* How the key AI feature 'Browse for Me' was not the initial concept, but came to be added last minute* Josh's initial skepticism about AI, up until just a few months ago, and what flipped him* Three strategies for AI integration* The perils of leading with consumer hooks, how that derailed Browser Co, and focusing on solving problems* Should cost be a key consideration when developing AI?Hope you enjoy us mixing it up this week by talking with someone wrestling with building in the space. Drop us a line on twitter at @nabeel and @fraser with what you want to hear more (or less) of.  (00:00) - Intro (01:02) - Welcome Josh, CEO of The Browser Company (02:51) - The origin of Arc Search (05:12) - Browse For Me (06:55) - Overview of Arc Search (16:51) - Thinking about competition (25:08) - User problems over elegant hooks (29:25) - Three strategies for AI integration (34:51) - The importance of prototyping (40:28) - The future of desktop search (46:01) - Thinking about costs when building with AI

Two former founders, now VCs, play with AI products and see where that leads...Fraser and Nabeel discuss the differences between horizontal disruption and vertical market disruption and the patterns of the phases of this disruption in the mobile age. From Adaptation, to Evolution, and eventually Revolution. Next, they dive into AI email app Shortwave AI. This leads to a conversation around what AI models are optimizing for, and how speed and polish can sometimes be part of the baseline usability of a product. They also explore the default Agent workflows we should be trying as we figure out the right knowledge worker AI copilot. Using the analogy of the history of web development and AI no code, from webflow to squarespace, what those design approaches might tell us about how AI workflow tools will develop.Finally, we discuss the potential value of talking to VCs when you're not raising, and the perils of taking even good generalized advice all the time. * Shortwave email* Avi Goldfarb and his book Prediction Machines* Avi also spoke with Patrick O'Shaughnessy * Yahoo Pipes wikipedia page, or see Retool's amazing history of Pipes* TLdraw's Makereal (00:00) - Intro (00:32) - "AI Week" at Startups (02:43) - Is it Adaption, Evolution, or Revolution (05:23) - Shortwave AI - Adapting email to AI (08:07) - Familiar but personalized (11:09) - Speed is a feature too (15:51) - Taking users on a behavior change journey (18:37) - What are Agent workflows we should default to (26:04) - TLdraw & Notion, over Pipes, as the core UX for AI (29:43) - TLDraw as metaphor for "show me the output" application building (31:13) - History of Web Development as analogy for AI no code (33:37) - Is it worth talking to VCs when you aren't raising (43:24) - Building relationships, Bizdev vs Sales

Reviewing voice note transcription apps, from WaveAI to Audiopen and ChatGPT. Then Fraser and Nabeel take stock of various ways AI coding is affecting software development, from Co-Pilots to "mealeable software development" and No-Code AI.  Lastly, the hosts talk through a wish list of products they'd love to see this year. Topics discussed:- Slowed Pace of AI Product Launches- Wave AI app, and what to look for in voice notes transcription- Current state of AI-enabled software tools- Potential impact of AI on software development workflows- What is the AI native storytelling medium- The AI todo list partner- Coding as a liberal art- User input layer and direct feedback in AI-enabled toolsReferences:- Wave AI - Meable Software in the age of LLMs- Zapier AI- Replit (00:00) - (01:39) - Wave AI (06:29) - Adapting Behavior to Leverage AI Products (12:54) - AI Enabled Software Development (13:46) - Infinitely Adaptable UIs via AI (18:31) - Zapier ++ (20:51) - AI Enabled Software Engineers (31:31) - Different Products Across Each Layer (34:48) - AI Products We'd Like to See in 2024

Fraser and Nabeel discuss the onslaught of AI hardware launches, exploring the potential factors behind this shift. They also discuss the importance of designing not by listening to your customer but by understanding the customer. The duo also discuss the importance and challenges surrounding 'prompt engineering' vs prompt writing, using BARD, SunoAI and Perplexity as case examples. They cover the layoffs in tech startups, examining the tension between growth and profitability and the potential drawbacks of a too-aggressive cost-cutting focus. Finally, they delve into the potential of AI in generating music, trying SunoAI.Links:- OpenAI Prompt Engineering guide - Rabbit - new AI hardware companion- Snipd - Podcast player- Remarkable - Dedicated tablet for note writing- SunoAI - AI music creation (00:00) - Designing AI Hardware, Prompt engineering vs prompt writing, Suno AI (00:47) - The era of AI hardware? (01:28) - Why it's the right time for AI Hardware startups (06:34) - YouTube, and building on top of another platform (08:37) - Is AI Hardware destined to be a phone appendage? (10:58) - Nobs, dials, and the joy of tactile (11:38) - The GPT Store: Should founders care? (17:57) - New year starts with a thud, layoffs (19:52) - How founders can pitch taking risk in 2024 (23:12) - Prompt Engineering Lives On (24:48) - Snipd's Summaries (27:37) - When do you use prompts vs your roll your own model (33:21) - My model vs yours, it depends on the goal (35:11) - Consumer Web Agents (37:25) - The problem isn't Actions, it's Context (43:25) - Exploring Bard and Gemini Pro (46:59) - Perplexity, Search and Synthesize (49:23) - Prompt Engineering vs Prompt Writing (52:32) - Suno AI (54:44) - What products lend themselves to Discord communities (56:25) - Live test of Suno AI

Nabeel and Fraser briefly discuss Google's new AI model, Gemini, and keeping authenticity in startup pitches. They discuss the perils of trying to simplify when technology is lending towards complexity. How to stay authentic to yourself when pitching and fundraising. The new findings of Claude prompting, and the likely continued need for prompt engineering. Finally, they talk about the startup SuperPowered, its potential pivot, and how passion for problem-solving can impact a company's direction.* Google Gemini's excellent product video "demo"* History of Razorfish * Claude 2.1 Prompting Technique* AI Meeting notes from Superpowered.me (00:00) - Building when entropy is increasing (01:00) - Introduction and Welcome (01:15) - Discussing the Frequency of AI Developments (01:45) - Google's AI Developments and the Gemini Team (02:19) - Explaining Gemini and its Significance (05:04) - Analyzing Google Gemini from afar (18:02) - You are 5 words away from being done (27:00) - The analogy of the early web vs early LLMs (30:05) - Do we need the Razorfish of AI (33:27) - The Future of AI Tools and Platforms (34:11) - The Importance of Implementation Engineers in AI (35:54) - Are we in an entropic or de-entropy phase? (37:57) - The Importance of Authenticity in Marketing (38:53) - Founders just being real (47:28) - How would you fundraise differently now? (50:15) - Superpowered.me and AI Note Taking Startups

Nabeel gets to ask Fraser for reflections on the development of ChatGPT on the one-year anniversary. They talk about the intersection of product managers and AI researchers, how ChatGPT came to be on the roadmap, how the team came to be formed, and the difference between polish & overengineering. They then discuss the recent explosion of real-time AI painting applications, the divide between text-based UIs and GUIs in AI products, and redesigning new creative arts at the intersections of art, realtime, and play.* Aliisa on the night of ChatGPT's launch* Martin Nebelong's Realtime painting experiements* Fastlane.AI Painting* Kepano - Photoshop for text(00:00) - Introduction (00:46) - Quotes from the ChatGPT 1 year anniversary (07:19) - Lessons from ChatGPT development (08:55) - It's not a happy accident (09:36) - The Birth of the AI Assistant (10:12) - Forming the ChatGPT team (19:15) - Differences between polish & overengineering (23:47) - Realtime AI Painting (25:47) - GUIs make it easier, so prompts can get harder (32:35) - Twitch meets the creative arts?

The conversation kicks off with the surprising news of Sam Altman’s departure from OpenAI. Nabeel and Fraser then discuss the ‘crazy week’ of product releases, demystifying the complexity and impact of AI in various workflows. They spotlight AI startup Lindy, review My GPTS, and early impressions of the product Dot, an AI-driven guide in daily life. The duo categorizes AI developments into three ‘buckets’: lead bullet strategy, a universal translator bucket, and inventing new behaviors. (00:00) - The three types of LLM products (02:07) - Exploring Lindy (08:21) - GPTs (15:52) - Frameworks for AI Product Development (20:21) - Dot: Your Guide in Life (26:07) - Revisiting Categorization (29:45) - Patience in Innovation