Peter Walker's LinkedIn Strategy, Head of…

Some personal news - today is somehow my last day at Carta. What an absolutely surreal experience. I joined just over 5 years ago with the fuzziest of mandates: go explore this private markets data,…

3.6K766538 viral

Career Updateabout 1 month ago

View on LinkedIn

Raising venture capital is HIGHLY dilutive. If you're a founder, how much of your own company will you retain over time? Let's break it down. All data from US startups on Carta (9,000+ in this study)…

LinkedIn post image: Raising venture capital is HIGHLY dilutive. If you're a founder, how much of your own company will you retain over time?

818957310 viral

Venture Capital2 months ago

View on LinkedIn

Does your city make the startup top 10? Okay, maybe not overall (this is based on dollars raised, so that's pretty tough to break into for smaller ecosystems). But how about in a specific vertical?…

LinkedIn post image: Does your city make the startup top 10?

5281626710 viral

Venture Capital2 months ago

View on LinkedIn

A basic recap of the AI story in 2026: 1. There has been an explosion of tokens as models have gotten more capable and agents are working autonomously (or nearly so). 2. As token usage has increa…

LinkedIn post image: A basic recap of the AI story in 2026:

628152379 viral

AIabout 1 month ago

View on LinkedIn

An illustration (with real data) of just how extreme the power law in VC can be: Data comes from 400+ US venture funds born in 2016, 2017, or 2018. The metric is net TVPI. 𝟮𝟱𝘁𝗵 𝗣𝗲𝗿𝗰𝗲𝗻𝘁𝗶�…

LinkedIn post image: An illustration (with real data) of just how extreme the power law in VC can be:

617154399 viral

Venture Capitalabout 2 months ago

View on LinkedIn

What are people actually paying AI to do right now? Collectively, users on OpenRouter spent 246 trillion tokens so far in August. Here's what they are paying for: 1. Workflow Execution - agents act…

LinkedIn post image: What are people actually paying AI to do right now?

70795488 viral

AI Agentsabout 1 month ago

View on LinkedIn

Topics & Content Focus

Primary Topics

AI Model Market Competition and Adoption EconomicsLLM Usage Intelligence Through Spend, Token, and Wallet-Share AnalysisFrontier AI Governance, Evaluation, and Industry Power DynamicsOpen-Weight Versus Proprietary Model Strategy

Secondary Themes

Model switching costs and ecosystem stickinessAI pricing, inference economics, and agentic consumptionThe influence of X on technology discoursePublic commentary on AI personalities and online communities

Industry Focus

Generative AI infrastructure and model platformsEnterprise AI procurement and adoptionAI developer ecosystems and API marketplacesFrontier AI policy, safety, and governance

Content Categories

Original market intelligenceData-led industry commentaryAI ecosystem analysisTechnology discourse observationSharp reactionary commentary

Performance Insights

STABLE

Performance Trend

Best Performing Topics

Provocative commentary on online behavior and AI tribalismAI market-share and adoption dataFrontier AI governance debates

Virality Signals

High comment volume relative to shares indicates debate and identification are stronger drivers than passive redistribution.Posts tied to recognizable AI labs, executives, and tribal conflicts have the greatest conversation potential.Data posts with a clear competitive narrative outperform purely descriptive analytics.Short, quotable observations can outperform substantially longer research-led posts.

Structure & Quality

HIGH

Depth Level

ADVANCED

Expertise Level

91/10

Uniqueness Score

Common Hooks

Contrarian comparisons that challenge the obvious definition of winningA striking market statistic followed by interpretationDirect commentary on a live controversy or public eventA short, provocative reaction to an online interaction

Common Endings

Open-ended strategic questionsA forward-looking uncertainty about what happens nextA concise emotional or humorous punchlineA recommendation to read, consider, or debate the source material

Value Delivery Methods

Transforms platform-level usage data into strategic conclusions about adoption and switching behavior.Separates token consumption from economic value rather than treating usage as a single metric.Adds second-order interpretation to product launches, pricing changes, and governance announcements.Connects technical-market developments with culture, media dynamics, and public incentives.

Formatting Style

Short paragraphs with generous whitespaceInline rankings and numbered questionsNarrative analysis supported by specific figuresConversational asides and occasional emojisMinimal use of hashtags and promotional formatting

Audience & Tone

YES

Question Usage

Detected Tone

Smart-casual expertData-literate and strategically mindedWry and socially observantConfident without being overly absoluteConversationally irreverentsemi-formalfirst-person

Interaction Style

Debate-orientedOpinionated and conversationalResponsive to controversy and status competitionInterested in interpretation beyond headline statistics

Community Building Signals

Creates shared context around rapidly changing AI industry events.Uses humor and cultural fluency to make expert commentary socially participatory.Frames analysis as an ongoing collective inquiry rather than a final verdict.Invites readers to compare competing explanations of market behavior.

Writing Style Patterns

Content Strategy

Hook: Contrarian comparisons that challenge the obvious definitionTone: semi-formalCTA: Implicit invitation to debate through unresolved q

Writing style breakdown

End of 2024 - the three big American labs (Anthropic, OpenAI, Google) held 87% of all dollars spent on LLMs on OpenRouter. And 69% of tokens used to boot. Just dominant.

Then their combined share of tokens starts to erode, beginning in May of 2025. It turns sharply downward as open weight models become a) better and b) stay much cheaper. That token share is down to 27% in recent weeks.

But the spend...they kept the spend. The three named labs still held 85% of all spend on platform to end 2025 and even now they hold 68 cents on every dollar.

Peter Walker

Warm Analysis

Performance Overview

Top Posts by Engagement

Posting Patterns & Frequency

Best Performing Days

Best Performing Times To Post