The Custom Everything Era with Kevin Rose

Summary of The Custom Everything Era with Kevin Rose

by Chris Hutchins

1h 7m•April 22, 2026
Listen to Episode

Overview of The Custom Everything Era with Kevin Rose

This episode of All The Hacks (host Chris Hutchins) features technologist Kevin Rose discussing how the latest consumer AI — personal agents, multimodal models, and no-code/low-code tools — are rapidly replacing traditional apps and workflows. They cover practical use cases (health tracking, photo/video creation, automations), privacy and security trade-offs, the coming crisis of authenticity in media, economic/job implications, and how these changes could free up more time for family and meaningful activities.

Key topics discussed

  • The rise of personal AI agents that replace single-purpose apps by building custom workflows on demand (no coding required).
  • Dramatic cadence of AI improvement: capabilities changing in weeks, not months.
  • Concrete consumer use cases showing time savings and new possibilities:
    • Protein/macros tracker built inside Claude Cowork as a persistent project.
    • Automated blood-work analysis that reads PDFs, produces charts, and cross-references with genome data.
    • Generating and sending custom images to a Samsung Frame TV via a small AI skill.
    • Curating best family photos, creating slideshows or video montages from texts and photos.
    • Automating price scrapes/updates for online gift-card sales.
  • Safety, trust, and privacy implications of uploading sensitive data (genome, health records, messages) to AI systems.
  • Security risks: third‑party packages and local code execution can be vectors for compromise; granting “computer use” permissions is powerful and potentially dangerous.
  • Authenticity crisis: AI-generated podcasts, videos and viral influencers will make it hard to trust online content; possible need for cryptographic provenance (e.g., camera-signed images).
  • Economic and job impacts: SaaS disruption, automation of many roles, but potential for new “artisan” or branded human-made goods and experiences.
  • Travel and parenting tangent: advice for visiting Tokyo with kids and why travel fosters curiosity in a fast-changing world.

Notable examples & anecdotes

  • Kevin built a personal macros/protein tracker in Claude Cowork that:
    • Lives as a persistent project
    • Accepts photos or plain-text meal notes
    • Writes daily nutrient-summary files to his desktop — replacing a paid app.
  • Blood-work + whole-genome upload: Claude detected a slightly high iron value and auto-checked genetic predisposition, surfacing actionable advice (give blood) without prompting.
  • Photo scraping task: Claude Cowork fetched high-resolution property images from a website, renamed them, and saved them to Google Drive — saving hours of manual work.
  • Automated gift-card pricing: a simple agent refreshed supplier spreadsheets/websites every 30 minutes and updated the e-commerce site to allow near-real-time listings.
  • Interview fraud examples:
    • Candidates using real-time AI to feed answers during technical interviews.
    • Spammers using AI on phone calls to sound human; a quick test (asking for a cupcake recipe) exposed the bot.

Notable quotes

  • “I just ask something and it goes out and it is in my personal agent that is quarantined to that aspect of my life, and it makes perfect software for me just the way I want.”
  • “We’re at the precipice of this moment of AI becoming aware enough and productive enough that it’s going to touch every single facet of our lives in the next couple of years.”

Main takeaways

  • Practicality: AI agents are already saving meaningful time on real tasks — but benefit often comes after you push the boundaries of what you ask them to do.
  • Rapid change: judge AI by what it can do today (or last week), not by past impressions; iteration speed is very high.
  • Personalization over apps: the “custom everything” era lets individuals define workflows and have AI build the software/automation for them.
  • Trust & authenticity are major emerging problems: provenance systems and verifiable channels (signed images, verified calls/videos) will become essential.
  • Security is nuanced: big vendors tend to have tighter access controls, but local execution, third-party packages, and granted permissions (like “computer use”) create risk.
  • Economic outcome uncertain: some industries will become more capital-efficient and see layoffs; others (brands, artisanal human-made goods, high-touch services) may retain value.

Risks & privacy/security considerations

  • Data sensitivity: be intentional about what you upload — genome and bloodwork are useful but sensitive. Chat transcripts and private messages may be the most intimate; many people choose not to share them.
  • Permissions to avoid: don’t give agents 2FA or full control of critical authentication tokens; use separate emails for financial/critical accounts.
  • Local-execution risks: installing third-party packages or enabling tools that control your machine can introduce malware/rogue code — exercise caution.
  • “Computer use” features: agents that can control apps, move your mouse, and interact with your files are powerful but can exfiltrate data if misconfigured.
  • Vendor trust: major models (OpenAI, Anthropic, Google) typically have enterprise-grade safeguards, but no system is risk-free.

Practical recommendations (what a listener can do next)

  • Small experiments:
    • Pick one repetitive task (e.g., daily nutrition logs, photo curation, price monitoring) and ask an AI to automate it.
    • Use a dedicated account or device for AI agents if you plan to give broad permissions.
  • Isolation & hygiene:
    • Use a separate email for high‑value accounts (banking, rewards).
    • Keep 2FA separate from any agent access.
    • Prefer disk encryption and limit granting “full disk access” to apps you trust.
  • Validate provenance:
    • Push for tools and standards that embed verifiable provenance into media (watermarks/signatures).
  • Learn by doing:
    • Spend time weekly experimenting — AI is evolving fast; staying hands-on prevents being left behind.
  • If you run a business:
    • Audit SaaS vendors for defensibility (brand, consumables/services).
    • Look for quick wins to automate operations with agents (e.g., pricing automation, content drafts, report aggregation).

Implications for creators, companies, and jobs

  • Creators: expect a flood of synthetic content; originality, authenticity, and verified provenance will be differentiators.
  • Companies / SaaS: many narrow SaaS products risk commoditization as agents and low-code tools replicate features quickly.
  • Jobs: automation will remove some lines of work but likely create new roles; historical precedent suggests new industries emerge, though scale and transition are uncertain.
  • Consumer preferences: “artisan” or branded real-human experiences and goods may increase in value as synthetic content becomes ubiquitous.

Tools & names mentioned

  • Anthropic (Claude, Claude Cowork, Claude Code)
  • OpenAI / ChatGPT
  • Google Gemini (Gemini Nano)
  • 11 Labs (voice)
  • Cloud/Cowork / CloudCode (variants of Claude tooling)
  • Leica (example for trusted photo provenance)
  • Examples of use-case platforms: Samsung Frame TV, Google Drive.

Where to follow Kevin Rose

  • The Kevin Rose Show (relaunch focused on AI; weekly, with plans for live streams to X and YouTube). Google “The Kevin Rose Show” to find the podcast and upcoming live schedule.

Final thought

Kevin and Chris argue the honest endgame many hope for isn’t less human time but more — AI could free time for family, creativity, and travel if we manage privacy, authenticity, and economic transition carefully. The immediate practical advice is simple: experiment, isolate sensitive data, and start automating small, concrete tasks today.