TickerTain
TickerTain
NewsroomShortsPortfolioConvergence
NewsroomShortsPortfolioConvergence
$ANTHROPIC·$MA····$INTC····$BLUE-ORIGIN·$SPCX····$CRWV····$CRM····$MSFT····$NVDA····$ORCL····$CURSOR·$AAPL····$OPENAI·$AMZN····$UBER····$GOOGL····$META····$TSLA····$DATABRICKS·$PERPLEXITY·$LYFT····$NBIS····$TSM····$LITE····$ANDURIL·
$ANTHROPIC·$MA····$INTC····$BLUE-ORIGIN·$SPCX····$CRWV····$CRM····$MSFT····$NVDA····$ORCL····$CURSOR·$AAPL····$OPENAI·$AMZN····$UBER····$GOOGL····$META····$TSLA····$DATABRICKS·$PERPLEXITY·$LYFT····$NBIS····$TSM····$LITE····$ANDURIL·
←

katelyn lesse

T3 · host / generalist

Joined Anthropic ~1 year ago from Stripe where she worked on Stripe Connect and the V2 Accounts API. Leads platform product and infrastructure for Anthropic's developer platform including managed agents, sandboxes, and MCP integrations.

1 call·1 name·100% bull·last heard last month·Kleiner Perkins
track record

no scored calls yet — needs a stated position or a categorical verdict, with a matured window vs SPY

top calls

highest conviction · one per company
1sthigh conviction
$ANTHROPICAnthropic

Anthropic reveals token economics strategy: advising, grading, dreaming patterns beat brute force

Anthropic is developing multi-model token strategies where smaller models execute with larger models advising, grading, and dreaming — achieving near-Opus performance at lower cost by making tokens non-fungible and assigning them specialized jobs.

Kleiner Perkins2026-08episode →

most discussed · click a bar to filter

  • $ANTHROPIC

recurring themes

  • AI Economics & Business Models2
  • AI Infrastructure2
  • Developer Tools2
  • AI Agents1
1 total
$ANTHROPIC
Anthropic
HIGHkatelyn lesse·Kleiner Perkins·last month·Inside Anthropic's 200-Person Platform Team | Katelyn Lesse and Angela Jiang
Anthropic reveals token economics strategy: advising, grading, dreaming patterns beat brute force
Anthropic is developing multi-model token strategies where smaller models execute with larger models advising, grading, and dreaming — achieving near-Opus performance at lower cost by making tokens non-fungible and assigning them specialized jobs.
"We've seen like sonnet executing with opus advising ends up getting almost opus level performance and it's actually cheaper than just sonnet because opus taught it how to do its j…"
21:30
9
AI Agentstailwind
Long-running autonomous agents replace human-in-the-loop workflows across knowledge work
Models now work for extended periods without human intervention, enabling agents to autonomously complete complex tasks like financial modeling (DCF analysis) with self-correction, shifting AI from chat assistants to independent knowledge workers.
9
AI Economics & Business Modelstailwind
Token economics shift from fungible commodities to specialized roles (advising, grading, dreaming)
Anthropic is pioneering non-fungible token strategies where tokens perform distinct jobs — advising, grading, dreaming — achieving higher intelligence-per-dollar than brute-force execution, moving beyond simple token pricing models.
8
AI Infrastructuretailwind
Agent infrastructure requires durable harnesses with ephemeral sandboxes for safety
The winning agent architecture separates a durable control plane (harness/brain) from ephemeral execution environments (sandboxes), enabling long-running agents that survive infrastructure failures while maintaining security boundaries.
8
AI Infrastructuretailwind
Modular agent architecture with durable harnesses and ephemeral sandboxes solves infrastructure bottleneck
The core constraint for long-running agents is infrastructure reliability — Anthropic's solution separates durable agent brains from ephemeral execution sandboxes, allowing agent recovery from failures without losing state, making infrastructure a commoditized layer.
8
AI Economics & Business Modelstailwind
Outcome-based token routing beats per-token pricing: Sonnet executing with Opus advising achieves near-Opus performance at lower cost
Token fungibility is a fallacy; assigning tokens different jobs (executing, advising, grading, dreaming) enables intelligence-per-dollar optimization — e.g., smaller model executes while larger model advises only on hard steps, reducing total tokens needed.
7
Developer Toolstailwind
Platform strategy: opinionated architecture, unopinionated infrastructure deployment
Anthropic aims to own the agent architecture layer (harness patterns, safety classifiers, MCP) while allowing customers to self-host infrastructure (sandboxes, MCP tunnels) — maximizing control for enterprises while maintaining architectural coherence.
7
Developer Toolstailwind
Harness engineering commoditized; differentiation moves to strategy layer (multi-agent coordination, token job assignment)
Basic agent loops (while loops with tool calling) are now well-understood; competitive advantage shifts to composing strategies — coordinating specialized agents, assigning tokens to advising/grading/dreaming roles — which is domain-specific and creates performance alpha.