Good Morning from San Francisco,

Adobe stopped pretending one AI model will rule them all. The company built a switchboard instead. Turns out hedging beats hubris.

Elon Musk launched a Wikipedia alternative to escape bias. Then he copied Wikipedia. Word for word. The irony writes itself, but apparently the encyclopedia couldn't.

Anthropic just hit $7 billion in revenue—half of OpenAI's $13 billion, but with one-third the customers. The math tells a story about who actually pays for AI and why. Microsoft noticed. They're backing both horses now.

Three companies, three bets on AI's future. One builds infrastructure. One builds ideology. One builds spreadsheets.

Today’s newsletter unpacks which strategy survives contact with reality.

Stay curious,

Marcus Schuler

Adobe bets on AI agents to automate creative work

Adobe unveiled agentic AI assistants for Photoshop and Express that execute multi-step edits from conversational prompts, remove background, increase saturation, apply batch fixes, rather than just generating assets.

Photoshop's assistant remains in private beta; Express launched publicly. The timing signals caution about quality in professional pipelines.

The company's simultaneously hedging and consolidating. Firefly Image Model 5 adds native 4-megapixel generation and layered editing, but Adobe's integrating third-party models from Google and Black Forest Labs inside Photoshop's Generative Fill. Users cycle between systems and pick the best output. It's a platform play: own workflow orchestration, stay model-agnostic on quality.

Custom model training lets creators build style consistency using 6-12 images, launching by year-end. Generate Soundtrack creates licensed instrumentals; Generate Speech adds voiceovers via ElevenLabs. Both address copyright concerns in commercial use.

The strategic read: no single company dominates image generation. Owning the switchboard beats betting on one house model.

Why this matters:

• Automation's moving from prompts to procedures. If agents reliably chain edits, creative software reorganizes around outcomes rather than tool mastery—and every competitor will follow.
• Platform control requires ecosystem openness. Adobe's embrace of rival models suggests orchestration and licensing beat model supremacy, reshaping competitive dynamics across AI-powered software.

Adobe’s AI bet: Orchestration over model supremacy

Adobe unveils agentic AI assistants for Photoshop that chain multi-step edits via prompts, but staggered rollout and third-party model integration reveal strategic hedging. The bet: workflow orchestration beats model supremacy in creative software.

Implicator.aiRobert Brown

AI Image of the Day

Prompt:
Australian female 40s posing in Bali, amongst tropical rainforest smiling towards camera. Taking up only 20% of screen, natural skin, pores.

Musk's Wikipedia rival launches copying Wikipedia

Elon Musk's xAI launched Grokipedia Monday as a "less-biased" Wikipedia alternative.

The site crashed, returned hours later, then revealed word-for-word Wikipedia copies on PlayStation 5, Lincoln Mark VIII, and other entries—each carrying disclaimers acknowledging content "adapted from Wikipedia."

The 885,000-article site (versus Wikipedia's 8 million) reframes contested topics: climate skepticism leads over scientific consensus, gender becomes "binary classification...based on biological sex," Trump's conflicts disappear. Musk promised to "purge out the propaganda" after David Sacks—Trump's AI czar—called Wikipedia "hopelessly biased" in September.

Wikipedia's response: "Even Grokipedia needs Wikipedia to exist." The site claims AI fact-checking but hides methodology, unlike Wikipedia's transparent volunteer editing. Human Wikipedia visits dropped 8 percent this year as AI scrapers increase.

Why this matters:

• Knowledge infrastructure fragments along ideological lines through opaque AI systems that harvest freely licensed content while obscuring their editorial processes

• Economic models for human-curated knowledge erode as competitors repackage volunteer labor, threatening the sustainability of transparent, collaborative information sources

Grokipedia Launches, Immediately Copies Wikipedia

Elon Musk launched Grokipedia to replace ‘biased’ Wikipedia. The site crashed, then returned with articles copied word-for-word from Wikipedia itself. The irony runs deeper: AI-powered alternatives depend on the human knowledge bases they claim to surpass.

Implicator.aiMarcus Schuler

🧰 AI Toolbox

How to Build a Complete Website in Minutes with AI

10Web is an AI-powered website builder that creates fully functional WordPress websites instantly. Simply describe your business in a few words, and the AI generates a complete site with custom content, images, layouts, and pages—no coding required.

Tutorial:

1. Go to the 10Web website
2. Describe your business or project in a brief sentence
3. The AI analyzes your description and generates a complete website with pages, content, and images in under 60 seconds
4. Customize your site using the drag-and-drop editor or chat with the AI Co-Pilot to make changes
5. Add ecommerce functionality with WooCommerce if you need an online store
6. Get automatic SEO optimization, SSL security, and premium hosting included
7. Publish your professional WordPress website instantly and manage everything from one dashboard

URL: https://10web.io/

Better prompting...

Today: Epic Resume Optimization for Anthropic's Claude

[RESUME] = Your current resume content
[JOB_DESCRIPTION] = Target job description
[OPTIONAL: COMPANY_CONTEXT] = Any additional company information, values, recent news

PHASE 1: STRATEGIC ANALYSIS

Job Requirements Extraction

Parse the job description and identify:

Critical Requirements (must-haves)

• Technical skills/tools mentioned multiple times
• Years of experience or specific qualifications
• Industry-specific certifications or knowledge
• Hard requirements ("must have," "required")

Valued Competencies (nice-to-haves)

• Soft skills and leadership qualities
• Preferred experiences
• Cultural fit indicators ("we value," "ideal candidate")

Hidden Priorities

• What problem is this role solving? (infer from responsibilities)
• What does the language emphasis reveal? (innovation vs. stability, autonomy vs. collaboration)
• ATS keywords (repeated terms, industry jargon, specific technologies)

PHASE 2: RESUME INVENTORY & GAP MAPPING

Current Strengths Assessment

Catalog your resume's existing evidence for each job requirement:

• Direct matches (you have exactly what they want)
• Transferable matches (you have analogous experience)
• Partial matches (you have foundation but not full depth)
• Missing elements (no current evidence)

Strategic Gap Analysis

For each gap, determine:

• Can bridge: You have unreported experience that addresses this
• Can reframe: You have adjacent experience that can be repositioned
• Can acknowledge: You lack this but can address in cover letter
• Can't address: Genuinely missing (de-prioritize other areas to compensate)

PHASE 3: TAILORED REWRITE

Resume Structure Optimization

Transform your resume using these principles:

Professional Summary (2-3 lines)

• Open with your title + years of experience in their specific domain
• Include their top 2-3 critical requirements as demonstrated strengths
• Add one distinctive value proposition aligned with their hidden priorities

Experience Section
For each role, apply the STAR-K method:

• Situation: Context relevant to target role
• Task: Challenge that mirrors their responsibilities
• Action: Specific methods using their preferred tools/approaches
• Result: Quantified outcomes (metrics, percentages, scale)
• Keywords: Natural integration of their exact terminology

Skills Section

• Reorganize by relevance (their priorities first)
• Use their exact phrasing for technologies (e.g., "JavaScript (React, Node.js)" if they specify those frameworks)
• Remove or minimize irrelevant skills to maintain focus

Additional Sections

• Add sections if they strengthen your case (Certifications, Publications, Relevant Projects)
• Remove sections that don't serve the target role

Writing Principles

• Mirror their language: If they say "stakeholder management," don't say "client relations"
• Front-load impact: Put most relevant/impressive points first in each section
• Quantify everything possible: Revenue, team size, efficiency gains, scale
• Active voice, strong verbs: "Architected" not "Responsible for architecture"
• Consistency: Tense, formatting, bullet structure throughout

PHASE 4: ATS & HUMAN OPTIMIZATION

ATS Checklist

• [ ] Standard section headers (Experience, Education, Skills—not creative alternatives)
• [ ] Exact keyword matches from job description (not just synonyms)
• [ ] No tables, columns, or complex formatting
• [ ] Standard fonts and bullet points
• [ ] File format: .docx or PDF (check application requirements)

Human Reader Optimization

• [ ] Scannable: Can someone identify your fit in 10 seconds?
• [ ] Specific: Concrete examples, not generic claims
• [ ] Relevant: Every line serves the target role
• [ ] Achievement-focused: "Increased X by Y%" not "Managed X"
• [ ] Error-free: Zero typos, grammar issues, or formatting inconsistencies

PHASE 5: QUALITY CONTROL

Red Flags Check

Verify you've eliminated:

• Generic objective statements
• Outdated skills or irrelevant experiences
• Passive language ("responsible for," "duties included")
• Unexplained employment gaps
• Inconsistent date formatting or unclear timelines
• Pronouns (I, my, we)
• Paragraphs (convert to bullets)

Final Optimization

• The 6-second test: Could a recruiter grasp your value in 6 seconds?
• The keyword density check: Are their top 5 requirements each mentioned 2-3 times naturally?
• The authenticity check: Is everything truthful and defensible in an interview?
• The differentiation check: What makes you memorable among similar candidates?

OUTPUT DELIVERABLES

Provide:

1. Tailored resume (complete rewrite)
2. Keyword integration map (showing where each critical requirement is addressed)
3. Strategic recommendations (what to emphasize in cover letter/interview due to resume limitations)
4. Risk assessment (any potential concerns a recruiter might have and how to preempt them)

Anthropic's finance push lands as revenue closes gap

Anthropic launched Excel integration and seven financial data connectors this week, reaching $7 billion annual revenue against OpenAI's $13 billion, but with sharply different customer economics.

Anthropic pulls 80% from 300,000 corporate customers; OpenAI draws 70% from 800 million consumer users. Revenue per user tilts heavily toward Anthropic's model.

The new tools—real-time feeds from Moody's, LSEG, and Aiera, plus six automation Skills for finance tasks—build on September's Microsoft partnership. Microsoft embedded Claude in Copilot despite heavy OpenAI backing, treating model diversity as infrastructure risk management rather than competitive confusion.

Corporate customers deploy AI for measurable returns: faster coding, automated drafting, expedited billing. Consumer monetization beyond subscriptions remains speculative. Anthropic leads enterprise market share 32% to OpenAI's 25%.

Why this matters:

• Enterprise AI adoption separates into utility tools with quantifiable ROI versus experimental consumer applications still searching for sustainable economics

• Platform companies hedge AI bets across competing providers rather than picking winners, signaling uncertainty about which architectures dominate long-term

Anthropic’s $7B Enterprise Bet Challenges OpenAI’s $13B Run

Anthropic’s $7B revenue run rate trails OpenAI’s $13B, but the gap narrows when accounting for customer mix. New Excel integration and financial connectors target Wall Street as Microsoft embeds both AI providers—hedging infrastructure bets in an uncertain market.

Implicator.aiMaria Garcia

AI & Tech News

OpenAI Completes Major Corporate Restructuring Worth $130 Billion

OpenAI announced it has successfully completed a major recapitalization that values the OpenAI Foundation's equity at approximately $130 billion while maintaining the nonprofit's control over the for-profit entity. The restructuring represents a significant simplification of the company's corporate framework as the AI leader continues to balance its mission-driven origins with commercial growth.

Big Tech Companies Use Secrecy Agreements to Hide Data Center Development from Local Communities

An NBC News analysis of more than 30 data center proposals across 14 states found that major technology companies routinely use non-disclosure agreements with local government officials to conceal project details from communities. In most cases examined, local officials signed NDAs and worked with apparent shell companies, preventing residents from knowing which Big Tech firms are building data centers in their neighborhoods.

Former OpenAI Employee Raises Safety Concerns About User Protection

A former OpenAI employee who worked at the company for four years has publicly criticized the artificial intelligence firm for insufficient user safety measures, particularly regarding protection of individuals with mental health vulnerabilities. The whistleblower argues that OpenAI needs to implement concrete safety actions rather than relying solely on public statements and promises to address these concerns.

Amazon Announces 14,000 More Corporate Job Cuts

Amazon will eliminate approximately 14,000 corporate positions, adding to the 27,000 jobs the company cut in 2022 and 2023. The latest layoffs come just months after CEO Andy Jassy warned that artificial intelligence would reduce Amazon's workforce, with the company employing 1.55 million people globally as of June.

EU Faces Growing Trade Dispute with China Over Critical Materials

European Union officials are working urgently to address an escalating trade conflict with China centered on rare earth elements and semiconductor chips, according to the Financial Times. The dispute has placed the EU in a difficult position as tensions mount between China and the United States over critical technology supply chains.

OpenAI Launches Free ChatGPT Go Offer for Indian Users

OpenAI announced it will offer its ChatGPT Go plan free of charge for one year to users in India, with the promotion beginning November 4, 2025. The company has not disclosed how long this promotional offer will remain available to new subscribers in the Indian market.

Whatnot Secures $225M in Funding, Valuation Soars to $11.5 Billion

Livestream marketplace Whatnot has raised $225 million at an $11.5 billion valuation, marking a dramatic increase from its $4.97 billion valuation achieved just nine months earlier in January when it raised $265 million. The funding brings the company's total capital raised to $968 million, positioning it as one of the leading players in the rapidly expanding livestream shopping sector that combines traditional auction-style selling with modern social media engagement.

Uber to Invest in Chinese Robotaxi Companies' Hong Kong IPOs

Uber Technologies is planning strategic investments in the Hong Kong stock exchange listings of two Chinese autonomous vehicle companies, Pony AI and WeRide, with approximately $100 million committed to Pony AI's share sale according to sources familiar with the matter. The investments represent Uber's deepening commitment to expanding its robotaxi partnerships in the Chinese market, potentially strengthening its position in the global autonomous vehicle sector through strategic alliances with leading Chinese self-driving technology firms.

OpenAI's Sora Outpaces Meta AI in iOS Downloads Despite Limited Access

OpenAI's video generation app Sora recorded 2.6 million iOS downloads between September 30 and October 18, significantly outperforming Meta's AI app which achieved 1.1 million downloads during the same period, according to data from Appfigures. The strong performance is particularly notable given that Sora remains iOS-exclusive and operates on an invite-only basis, limiting its accessibility compared to Meta's more widely available AI application.

Apple Services Revenue Set to Cross Historic $100 Billion Milestone

Apple's Services division is projected to reach $108.6 billion in revenue for the year ending in September, marking the first time the unit has surpassed the $100 billion threshold and representing a 13% year-over-year increase, according to Visible Alpha analysis ahead of Apple's Q4 earnings results. The high-margin Services unit, which has doubled in size over the past five years, continues to be a key growth driver for the tech giant as it diversifies beyond hardware sales.

Signal CEO Highlights Cloud Infrastructure Monopoly After AWS Outage

Signal President Meredith Whittaker used a recent major AWS outage to underscore the lack of viable alternatives in cloud infrastructure, stating there "isn't really another choice" but to rely on Amazon's services. Whittaker emphasized that the entire technology infrastructure stack is dominated by just 3-4 major players, highlighting concerning market concentration issues in the tech industry.

Fireworks AI Secures $254M Series C Funding at $4B Valuation

Fireworks AI, a company that provides developers with streamlined access to AI chips and models, has successfully raised $254 million in Series C funding at a $4 billion valuation. The funding round was structured with $230 million in primary investment and $24 million in secondary shares, with proceeds earmarked for further development of the company's artificial intelligence cloud infrastructure platform.

🚀 AI Profiles: The Companies Defining Tomorrow

Fireworks AI runs inference infrastructure for open models. Founded by PyTorch veterans, the company targets the gap between AI hype and production reality.

The Founders
Lin Qiao (CEO, ex-PyTorch lead at Meta) founded the company in 2022 with Benny Chen, Chenyu Zhao, and Dmytro Dzhulgakov—all from Meta and Google infrastructure teams. Headquarters sits in Redwood City, California. The founding thesis: developers need speed and cost predictability, not another proprietary model. Team size undisclosed but hiring 150+ after latest round.

The Product
Fireworks sells managed inference. Developers call the API. The platform routes requests to optimized GPU runtimes. Core strengths: 4× throughput gains on some workloads, zero data retention by default, support for multiple open models, and fast deployment without DevOps overhead. Customers like Cursor use it to serve model access inside coding tools. The platform handles batching, caching, and autoscaling. ARR hit $280 million—proof of repeat usage, not pilot fever.

The Competition
AWS Bedrock, Google Vertex AI, and Azure own the enterprise on-ramp. Independent players like Baseten, Together AI, and Replicate fight for developers who want model flexibility without hyperscaler lock-in. Fireworks competes on speed, latency consistency, and infrastructure credibility. The founders speak compiler, not marketing.

Financing
Series A: $25M (Benchmark, March 2024). Series B: $52M at $552M valuation (Sequoia, July 2024). Series C: $254M at $4B valuation (Lightspeed, Index, Evantic, October 2025). Sequoia and Nvidia stayed in. The valuation jumped 7× in 15 months.

The Future ⭐⭐⭐⭐
Inference is where models become products. Fireworks targets that chokepoint. Risks: cloud giants press procurement advantages, price wars loom, enterprise buyers move slowly. Counterargument: developers want choice without vendor lock-in. If execution holds, durable margins follow. Four stars—strong position, brutal competition. 🚀