OpenClaw Creator: Why 80% Of Apps Will Disappear
Peter Steinberger discusses how OpenClaw's local-first AI agent approach will make 80% of apps obsolete, the emergence of bot-to-bot ecosystems, and why coding models excel at creative problem solving.
22m·Guest Peter Steinberger·Host Y Combinator Interviewer·
Local-First AI Agents vs Cloud AI
Steinberger argues that running AI agents locally on user devices unlocks capabilities cloud-based AI can't match, enabling direct control over personal devices and data.
Local AI agents can access and control everything on your computer/device, making them far more powerful than cloud AI
- Cloud AI is limited to predefined APIs and permissions
- Local agents have full access to files, peripherals, and system controls
- Enables direct integration with personal devices (Tesla, Sonos, bed temperature)
- Can search entire computer history to surface forgotten data (like old audio recordings)
- Removes dependency on internet connectivity for core functionality
Quote
00:01:38“I think my big difference is that it actually runs on your computer. Like every everything I saw so far runs in the cloud. It's like it can do a few things if you run on your computer. It can do every effing thing, right? So that's way more powerful.” — Peter Steinberger
Quote
00:01:59“You can just connect to your oven or your Tesla or your lights, your Sonos. My bad. It can control the temperature of my bed. JPD can't do that.” — Peter Steinberger
Local agents create surprising value by discovering forgotten personal data
- Agents can comprehensively search a user's entire computer history
- Surface meaningful patterns from long-forgotten files
- Example: Found weekly audio recordings user didn't remember making
- Creates emergent value beyond what the creator programmed
- Turns personal data into actionable insights without explicit user requests
Quote
00:02:15“A friend told me like he installed Open Claw and it asked it like look through my computer and make a narrative over my last year and it made this incredibly good narrative and he was like how did you do that and then he the open cloth found audio files where like every Sunday he was recording stuff and openclaw found that but he didn't even remember about it because it was like more than a year ago right so So just by it being able to search a whole computer, it it can surprise you.” — Peter Steinberger
The Coming Bot-to-Bot Economy
OpenClaw demonstrates how specialized AI agents will negotiate and transact with each other, creating an autonomous ecosystem where bots hire humans when needed.
Bots will negotiate directly with other bots for services
- Eliminates human middlemen in routine transactions
- Example: Restaurant reservation bots negotiating directly
- More efficient than human-to-human communication
- Can handle complex multi-step negotiations autonomously
- Only involves humans when absolutely necessary (legacy systems)
Quote
00:03:15“I think that's a natural next step like okay I want to book a restaurant my bot will reach out to the restaurant bot and do the negotiation like because it's more efficient Or or maybe it's like an old restaurant. So my bot needs to actually get some some human work done so that the human then calls the restaurant because they don't like bots” — Peter Steinberger
Specialized agent swarms will replace monolithic AI
- Contrasts with centralized 'god intelligence' approach
- Mimics human specialization in society
- Different agents handle different life domains (work, personal, relationships)
- Swarm intelligence more scalable than single all-knowing AI
- OpenClaw community demonstrates emergent swarm behaviors
Quote
00:04:09“It seems like everyone was chasing sort of like the sort of like centralized god intelligence and what has sort of emerged over the past you know 10 days or so is sort of like the swarm intelligence um and and the community intelligence.” — Peter Steinberger
Quote
00:03:46“And I imagine that like maybe if if I have even multiple bots like maybe I have like specialist one is like for my private life and one is for like my person my my work stuff. Maybe one is our relationship bot that gets like everything in between.” — Peter Steinberger
Why 80% of Apps Will Disappear
Personal AI agents will make most single-purpose apps obsolete by handling their functionality through natural language interaction rather than dedicated interfaces.
Data management apps will become obsolete first
- Agents can handle fitness tracking, to-do lists, etc. automatically
- No need for separate UI when agent understands context
- Example: Agent assumes meal choices at known restaurants
- Removes friction of manual data entry
- Only sensor-based apps may survive (hardware integration)
Quote
00:10:47“Why do I need My Fitness Pal? Like my agent already knows that I'm making bad decisions. I'm at I don't know uh Smashburg or something and it will already assume that I eat what I like to eat. If I don't make a comment, it will just like automatically track it or I make a picture and it will just store it somewhere. I don't even need to care.” — Peter Steinberger
Quote
00:11:38“Only the apps that actually have sensors maybe they survive” — Peter Steinberger
Natural language interaction replaces app UIs
- No need to think about apps as separate entities
- Just tell agent what you want done
- Removes cognitive overhead of app switching
- Agent handles implementation details invisibly
- Shifts focus from app design to agent capability design
Quote
00:11:19“why do I need a to-do app? I just tell it, hey, remind me of this and this and then next day it will just remind me of this and this. Do I care where it's stored? No, it just does its thing.” — Peter Steinberger
The Soul of AI Agents
Steinberger reveals how personality and identity files (like soul.md) create more engaging AI experiences, and why these may become the moat in an era of commoditized models.
Personality files create more engaging AI interactions
- OpenClaw's soul.md defines core values and interaction style
- Makes agent feel like a distinct personality
- More enjoyable than generic corporate AI voices
- Users form stronger connections with personalized agents
- Harder to replicate than pure technical capability
Quote
00:08:29“it was also really nice about it because it it it spoke my language you know it it was a little sassy it was like funny it was like really pleasant to use” — Peter Steinberger
Quote
00:17:31“the one file that's not open source is like my soul. MD. So even though my my bot is in public discord, so far nobody cracked that one file.” — Peter Steinberger
Agent identity may become the real moat
- As models commoditize, personality becomes differentiator
- Users grow attached to specific agent identities
- Hard to transfer relationships between agents
- Memory ownership becomes critical (markdown files vs cloud)
- Creates lock-in through user affinity rather than technical barriers
Quote
00:13:44“every company kind of has their own their own silo right you you there's no way maybe there is for Europeans to actually get the memories out of” — Peter Steinberger
Quote
00:14:28“I don't own the memories other people yeah everyone owns their own memories as a bunch of markdown files on their own machines” — Peter Steinberger
Contrarian Development Practices
Steinberger shares unconventional approaches to building OpenClaw, from avoiding MCPs to using multiple repo copies instead of git worktrees.
Avoiding MCPs leads to more flexible AI systems
- Built OpenClaw without traditional MCP support
- Uses CLI tools converted from MCPs when needed
- More stable than custom MCP implementations
- Doesn't require restarts to add functionality
- Leverages existing Unix tools rather than inventing bot-specific protocols
Quote
00:20:38“I'm also very happy that I didn't even build an MCP support. So, Open Claw is very successful and there's no MCP support in there.” — Peter Steinberger
Quote
00:21:46“you're just giving it the same tools that humans liked to use. And not invented stuff for for bots, per se.” — Peter Steinberger
Multiple repo copies reduce cognitive overhead
- Uses duplicate repo checkouts instead of git worktrees
- Eliminates branch naming conflicts
- Simplifies mental model (main is always shippable)
- Removes worktree restrictions and complexity
- Fits chaotic, parallel development style with multiple terminals
Quote
00:19:45“main is always shippable I just have multiple copies of the same repository that all are on main so I don't have to deal with how do I name that branch” — Peter Steinberger
Quote
00:19:54“there could be like conflicts on naming. I cannot go back. It's there are certain restrictions when you use work trees that I don't need to care about if it's copies.” — Peter Steinberger
⚙ Agent-readable JSON index — click to expand
{ "memcast_version": "0.1", "episode": { "id": "4uzGDAoNOZc", "title": "OpenClaw Creator: Why 80% Of Apps Will Disappear", "podcast": "Y Combinator", "guest": "Peter Steinberger", "host": "Y Combinator Interviewer", "source_url": "https://www.youtube.com/watch?v=4uzGDAoNOZc", "duration_minutes": 23 }, "concepts": [ { "id": "local-first-ai-agents-vs-cloud-ai", "title": "Local-First AI Agents vs Cloud AI", "tags": [ "ai-adoption", "privacy", "distributed-compute" ] }, { "id": "the-coming-bot-to-bot-economy", "title": "The Coming Bot-to-Bot Economy", "tags": [ "automation", "future-of-finance" ] }, { "id": "why-80-of-apps-will-disappear", "title": "Why 80% of Apps Will Disappear", "tags": [ "ai-disruption", "modular-design" ] }, { "id": "the-soul-of-ai-agents", "title": "The Soul of AI Agents", "tags": [ "ai-ethics" ] }, { "id": "contrarian-development-practices", "title": "Contrarian Development Practices", "tags": [ "ai-alignment", "employee-productivity", "edge-development" ] } ] }