Skip to content
Aakash GuptaAakash Gupta

These 3 AI Browsers make Chrome Feel useless

Naman Pandey breaks down AI agent browsers head-to-head. ChatGPT Atlas vs Perplexity Comet vs Arc Dia. Live demos, exact workflows, and which browser wins for each PM use case. Full Writeup: https://www.news.aakashg.com/p/naman-pandey-podcast --- Timestamps: 0:00 - Intro 1:53 - Overview of All Three Browsers 3:00 - Installing the Browsers 3:39 - Demo: Tab Context for Research 6:01 - Atlas: Auto-Filling Job Applications 8:08 - Atlas: LinkedIn Scraping for Outreach 10:46 - Ads 12:43 - Atlas: Gmail Integration 17:12 - Productivity Use Cases for PMs 17:34 - Perplexity Comet: Research & Shopping 27:07 - Dia: YouTube Video Summarization 30:17 - Ads 33:03 - Dia: Jira & Atlassian Integration 38:39 - Weaknesses of Each Browser 43:27 - Building the Use Case Mind Map 48:41 - Final Rankings: Which Browser Wins 53:39 - Advanced Tips & Tricks 55:55 - Outro --- 🏆 Thanks to our sponsors: 1. Jira Product Discovery: Plan with purpose, ship with confidence - https://www.atlassian.com/software/jira/product-discovery 2. Mobbin: Discover real-world design inspiration - http://mobbin.com/aakash 3. Pendo: the #1 Software Experience Management Platform - http://www.pendo.com/aakash 4. Product Faculty: Get $550 off the AI PM Certification with code AAKASH550C7 - https://maven.com/product-faculty/ai-product-management-certification?promoCode=AAKASH550C7 5. Land PM job: 12-week experience to master getting a PM job - https://www.landpmjob.com/ --- Key Takeaways: 1. AI agent browsers are underhyped for PMs - Only 2 out of 500 PMs at Berkeley were using them. If you're doing web research, competitor analysis, or data scraping, you're leaving hours on the table every week. 2. The three browsers serve different purposes - ChatGPT Atlas for deep research across multiple pages. Perplexity Comet for real-time quick lookups. Arc Dia for workflow automation. They're not competing head-to-head. 3. Atlas dominates data extraction - Scrape YC companies, find recruiters on LinkedIn, build competitor comparison tables. What took 2-3 hours now takes 10 minutes with one prompt. 4. Comet wins on speed for real-time info - Stock prices, sports scores, breaking news. It's the fastest by far. Perfect for quick research sprints across Reddit, Twitter, and news sites. 5. Dia automates repeated workflows - Monitor competitor pricing weekly. Document onboarding flows. Generate recurring reports. Set it once, let it run on schedule. 6. Tab context is the hidden superpower - Open 5 competitor sites. Ask "What's the common pricing strategy?" The AI reads all tabs and synthesizes insights. Eliminates copy-paste friction. 7. The job seeker use case is mind-blowing - "Find 20 PMs at Google, get their LinkedIn profiles, draft personalized DMs." Atlas does this in 15 minutes. Used to take 2-3 hours manually. 8. Onboarding analysis becomes trivial - "Go through Notion's signup flow, capture screenshots, document each step." Dia does this in 5-10 minutes. Perfect for competitive analysis. 9. Don't log into sensitive accounts - Banking, email, social media with private data - keep these in your regular browser. Use AI browsers only for public research and data extraction. 10. The slowness matters less than you think - Yes, they're slow compared to Google. But if the alternative is 2 hours of manual work, waiting 10 minutes is a massive win. Batch requests and walk away. --- 👨‍💻 Where to find Naman Pandey: LinkedIn: https://www.linkedin.com/in/namanpandey0796/ Podcast: https://www.youtube.com/@ReadySetDoPodcast 👨‍💻 Where to find Aakash: Twitter: https://www.x.com/aakashg0 LinkedIn: https://www.linkedin.com/in/aagupta/ Newsletter: https://www.news.aakashg.com #aibrowsers #aipm --- 🧠 About Product Growth: The world's largest podcast focused solely on product + growth, with over 200K+ listeners. 🔔 Subscribe and turn on notifications to get more videos like this.

Naman PandeyguestAakash Guptahost
Jan 29, 202657mWatch on YouTube ↗

CHAPTERS

  1. 0:00 – 1:28

    Head-to-head setup: Perplexity Comet vs ChatGPT Atlas vs Dia

    Aakash introduces Naman and frames the episode: compare three AI browsers in real workflows to decide which is best for most people. Naman previews that they’ll cover strengths, weaknesses, and final rankings so viewers can pick based on use case.

    • Goal: compare the three AI browsers “head to head”
    • Audience focus: productivity boosts for product managers and general users
    • Promise of a clear recommendation at the end
  2. 1:28 – 2:14

    Quick 300-foot view: what each browser is best at

    Naman gives a high-level positioning for each tool. Comet shines for real-time research, Dia excels at remembering context across tabs, and Atlas is the most agentic for taking actions in the browser.

    • Comet: research-oriented tasks, especially real-time
    • Dia: strongest tab-as-OS context retention
    • Atlas: best for agentic actions (performing operations, not just summarizing)
  3. 2:14 – 2:53

    Getting set up: where to download and how installation works

    They walk through how to install each browser and what to expect during setup. The process is straightforward—download, then drag into Applications on Mac.

    • Comet download link and basic install flow
    • Atlas download accessed from within ChatGPT UI
    • Dia download via diabrowser.com and DMG install steps
  4. 2:53 – 5:07

    Baseline superpower: using open tabs as research context (Nvidia one-pager demo)

    Naman demonstrates a capability common to all three: consolidating info across open tabs into a structured deliverable. He uses multiple Nvidia-related pages and asks the browser to generate a succinct one-pager with graphics.

    • All three can read and synthesize across relevant open tabs
    • No copy/paste needed—browser fetches context directly
    • Auto-grouping relevant tabs even with unrelated tabs open
    • Useful for creators/PMs doing multi-tab research and writing
  5. 5:07 – 7:11

    Atlas agent demo: auto-filling job applications from your resume

    Atlas is showcased as an action-taking agent that can complete tedious workflows. Naman uploads a resume and instructs Atlas to fill an entire job application form, including generating narrative answers beyond what’s explicitly on the resume.

    • Upload resume → prompt: “Fill out this application for me”
    • Watches the agent fill fields live, end-to-end
    • Generates answers for open-ended questions (motivations/fit)
    • Reframes PM skill: supervise agents, multitask, return to results
  6. 7:11 – 13:20

    Atlas for outreach: LinkedIn discovery + “scraping” via agent mode behaviors

    Naman shows how Atlas can navigate LinkedIn to assemble a prospect list for podcast outreach, including links to profiles. He explains how it can click into “Contact info” areas to extract emails/phone numbers into a spreadsheet-like output—behavior that’s hard to replicate with traditional bots.

    • Agent can search LinkedIn, open profiles, compile candidate lists
    • Uses website context (podcast site) to infer guest criteria
    • Can click-through UI layers (e.g., hidden contact info)
    • Guardrails: avoids the term “scrape,” but works in agent mode with consent
  7. 13:20 – 16:48

    Atlas + Gmail integration: automated inbox analysis for recurring subscriptions

    They shift from generic login workflows to first-class integrations. With Gmail authorization, Atlas can scan emails to find recurring expenses and even link to customer support/cancellation paths, turning a painful manual audit into an automated report.

    • Built-in Gmail connector avoids manual login inside agent flow
    • Find recurring subscriptions by scanning receipts over long history
    • Outputs lists and identifies uncertain items for review
    • Broader theme: AI excels at scanning “1,000 things,” not just top results
  8. 16:48 – 20:50

    Pivot to Perplexity Comet: research and shopping comparison agent

    Naman positions Comet as the research workhorse and demonstrates a gift-finding task that compares prices across retailers, not just Amazon. The browser can then move results into Google Sheets and refine the output with more detailed prompts.

    • Prompt: gift ideas for a 10-year-old + compare prices outside Amazon
    • Finds cheaper listings on other sites (e.g., Barnes & Noble)
    • Can populate a Google Sheet tab with results
    • Handles iterative refinement: add links, exact prices, notes
  9. 20:50 – 24:17

    Comet’s edge: extensions, historical price data, and token-heavy workflows

    They explain how Comet gains extra power through extensions like Honey/Capital One Shopping to access historical price and discount data. Naman notes the surprisingly advanced capabilities available even on free tiers and comments on how token-intensive this must be behind the scenes.

    • Historical pricing comes via connected extensions, not native storage
    • Useful for deal-hunting: biggest discounts vs earlier in the year
    • Strong fit for time-sensitive, comparison-heavy research
    • Observation: complex operations appear available even for non-Pro users
  10. 24:17 – 26:23

    Reliability check: hallucinations, link accuracy, and Sheets as a PM “bread and butter”

    Aakash asks about hallucinations; Naman reports few issues and emphasizes that links and prices tend to be accurate. They broaden the takeaway: Comet + Sheets enables practical analysis like budgeting insights or reconciling PDFs with spreadsheet data.

    • Low hallucination observed in real browsing tasks
    • Links and prices validate accurately against sources
    • Sheets-centric workflows: expense tracking, planning, analysis
    • Combine PDFs + Sheets + open tabs for cross-source updates
  11. 26:23 – 32:17

    Dia’s experience and superpower: context memory across tabs (YouTube script demo)

    Naman praises Dia’s polish and onboarding, then demos using two YouTube tabs as context to draft a script with a strong hook. Dia performs especially well at determining which tabs are relevant—even when videos aren’t actively playing—making it powerful for synthesis across messy tab sets.

    • Best-in-class onboarding and product polish (PM case study)
    • Uses multiple YouTube tabs to draft a new video script
    • Strong at retaining/identifying relevant tab context automatically
    • Limitations appear in some integrations (e.g., Gmail requiring manual uploads)
  12. 32:17 – 38:27

    Dia for enterprise/PM workflows: Jira, GitHub, Loom → auto-generated tickets

    They cover Dia’s Atlassian-aligned integrations, even though Naman can’t fully demo them. Dia can translate GitHub issues or Loom bug videos into structured Jira tickets with details and transcripts, streamlining a common PM/engineering interface.

    • Direct Jira integration aimed at enterprise workflows
    • Generate Jira tickets from GitHub repo context and bug info
    • Use Loom video context to create transcripts + ticket details
    • Best fit if your org runs Atlassian Suite heavily
  13. 38:27 – 42:40

    Where each browser breaks: speed, navigation/CAPTCHA, dark patterns, and privacy tradeoffs

    Naman outlines practical constraints. Atlas can be slower for urgent tasks; Comet can struggle with long navigational chains or tricky UX hurdles; Dia raises concerns about controlling sensitive information exposure based on reports and its deep context approach.

    • Atlas: robust but slower for time-critical operations
    • Comet: can struggle with complex navigation chains/CAPTCHAs/dark patterns
    • Dia: harder to exclude private info; some reported leakage concerns
    • General guidance: choose based on privacy appetite and workflow risk
  14. 42:40 – 48:07

    Use-case mind map: PM vs general use cases, plus “don’t use it for…” list

    They collaboratively build a mental model of when AI browsers outperform standalone LLMs. PM use cases include competitive analysis, sentiment research, structured note-taking, documentation generation, and data analysis; general use cases include email triage, shopping comparisons, and scraping—while non-use cases include highly sensitive workflows and long, dark-patterned cancellation paths.

    • PM: competitive analysis, sentiment tracking, structured research notes
    • PM: documentation generation across flows (screenshots + step-by-step)
    • General: shopping companion, email scanning, scraping without coding
    • Avoid: sensitive logins, long navigation chains, subscription cancellations
  15. 48:07 – 57:31

    Final rankings + advanced advice: which browser wins and how to capture the ‘alpha’

    Naman ranks Atlas as best overall for most people due to agentic action and scraping, with Comet close behind for research, and Dia third for narrower flagship use cases. They discuss pricing/rate limits and conclude with advice: try features immediately when they ship because the advantage window won’t last.

    • Overall winner: ChatGPT Atlas (agentic operations, scraping utility)
    • Comet: best usability/clarifying questions; strongest research workflows
    • Dia: strong context and Atlassian niche, but narrower default value
    • Tips: watch release notes; test new features early; rate limits minimal for Plus users

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.