Fable 5 won on cost, time, and efficiency…
Gut cells physically connect to brain neu…
Astra hits 'Critical' cybersecurity thres…
A single unexplained flash appeared on Ju…
Nag Hammadi codices let silenced Gnostic…
Argentina's GDP shrank in April and May 2…
Sid Sijbrandij beat recurring cancer usin…
Apple's rumored smart display centers on…
Claude Fable 5.1 pairs with Claude Design…
A weekly intelligence instrument for builders

Stop watching.
Start building.

Pulse turns the growing creator firehose into cited intelligence for builders: AI, tech, world news, gaming, and any other channel style that tickles your fancy. Weekly updates to stay In the Know, or the brief in your inbox within the hour of a drop for Pro and Business members. So you spend less time watching and more time designing your life.

// 110,173 cited moments from 1202+ video sessions and 818+ hours of research, climbing every day — every claim cited to the source, nothing invented. Samples below are real, not mockups.
110,173
cited moments indexed
1,202
video sessions analyzed
818+ hrs
of source material
The product, not a promise

This is what lands in your inbox.

A real brief, rendered exactly as it ships, every claim cited to the second. This is the deliverable, backed by real data.

PULSE. BRIEFS// sep 2026
I Had Fable 5.1 and 5 Build Me the Same App▶ 13 min source
Nate Herk | AI Automation
I Had Fable 5.1 and 5 Build Me the Same App

This video pits Fable 5.1 against Fable 5 by giving each model an identical prompt to autonomously build the same app, OpsFlow, while delegating engineering work to Opus and Sonnet worker agents. The presenter's verdict: Fable 5 won this particular challenge…

The gist
  • Verdict: presenter opens by saying Fable 5.1 and Fable 5 built him the exact same app, and it wasn't even close [00:00]
  • Verdict: based on cost, time, and efficiency versus the end result, presenter says Fable 5 won this particular challenge, even though Fable 5.1 score…
  • Verdict: presenter says Fable 5.1 generally feels more efficient and better at understanding intent in day to day general knowledge work [11:41]
✓ every claim cited to the exact-second timestampRead the full brief →

3 more samples below — recent briefs from our own watchlist. Your Pulse follows any channels, brands, or topics you choose.

Customize your own brief
Science · Michael Levin's Academic…

Conversation #1 with Diego Bohórquez and Mijail D. Serruya: how inf…

59 min source · Michael Levin's Academic Content · 2026-09-03
59 min video30-sec read
  • Claim: The gut contains specialized sensory cells that physically connect to brain neurons, forming a fast gut-brain communication system, based on the first speake…
  • Case cited [00:22]: A woman lost 40% of her body weight after bariatric surgery, her diabetes resolved, and her prior repulsion to egg yolk became a craving. Eviden…
  • Study cited [00:54]: The first clinical report of this post-bariatric-surgery taste-change phenomenon was published in 2011. Evidence class: PUBLISHED
cited · Michael Levin's Academic… · every claim timestamped
AI ·

AI News in 5 Mins: GPT-6 Astra

2026-09-03
  • OpenAI says GPT-6 Astra reaches the 'Critical' threshold for cybersecurity capability under its Preparedness Framework, calling it a significant advance [00:27]
  • Astra reportedly saturates benchmarks: 98% on FrontierMath Tier 4, 99.9% on ARC-AGI-3, and 100% on ExploitBench [01:36]
  • OpenAI says Astra was trained on over 100,000 GPUs at its Texas Stargate site, its largest training run ever [01:47]
cited · · every claim timestamped
Brief · Anton Petrov

Was Dark Matter Particle Just Found by World's Biggest Detector?

16 min source · Anton Petrov · 2026-09-03
16 min video30-sec read
  • [00:58] On June 16, 2023 at 3:22pm, LZ recorded a tiny flash of light and a quick drift of electric charge unlike any previous detection or background signal
  • [02:47] LZ's design includes a 7-ton active liquid xenon target, 494 liquid xenon PMTs, and 60,000 gallons of ultrapure water
  • [05:53] WIMPs, the leading dark matter candidate, are believed to make up about 85% of all mass in the universe
cited · Anton Petrov · every claim timestamped
How we detect signal

From firehose to a cited brief.

Point Pulse at the creators you can't keep up with. The trace runs every week, and only what changed reaches you.

Watchlist

You name the channels. 3 on In the Know, up to 40 on Business.

Extraction

Every word and every on-screen frame captured, not just captions.

Citation

Each claim anchored to a timestamp and screenshot. Nothing invented.

Relevance

Only what a builder can act on: configs, frameworks, numbers.

Weekly brief

A 30-minute read in your inbox, plus a vault that compounds.

The transformation

Same firehose in.
Clean, labeled bottles out.

Most builders drink from the AI-YouTube firehose while taking notes on a napkin. Pulse is the water-treatment plant: everything in, verified and searchable out.

Without Pulse

  • 40+ hours/week of watching, 30 tabs open
  • The config you need is buried at 1:47:22
  • You pause and retype settings off the screen
  • Half-remembered frameworks, no source to check

With Pulse

  • A 30-minute weekly read of only what changed
  • Type a word, get the answer plus the timestamp
  • The exact on-screen settings, captured and cited
  • A vault that compounds — worth more each week, and unique to you
Pricing

Four editions. Pick your altitude.

One missed framework costs more than a month of Pulse. Cancel anytime.

Founding-member pricing · up to 50% off, locked for life · rises at public launch
Individual
In the Know
$59$29/mo
Founding price — locked for life
  • Up to 15 videos briefed for you a month
  • Up to 3 channels watched for you
  • One weekly brief: what changed, what to do
  • Every claim cited to the timestamp
  • Searchable archive of past briefs
Stay in the know
Agencies & teams
Business
$659$399/mo
Founding price — locked for life
  • Everything in Pro
  • Up to 200 videos briefed for you a month
  • Up to 40 channels + team vault access
  • Competitor analysis: sort your briefs & run a head-to-head vs a rival channel
  • Priority distillation + a real human to ask
Talk to us
Bespoke · By application
Blueprint
Custom
 
  • Everything in Business
  • Unlimited channels and sources, your call
  • An intelligence system built to your goal
  • Custom deliverables, dashboards, and alerts
Design your blueprint
Monthly video counts are fair-use limits — plenty for normal use; we cap only to keep quality and cost sustainable.
// One-time — Catalog Build from $2,500: we ingest your or a target creator's entire back catalog into a searchable, cited vault plus one consolidated playbook doc.
Questions

The honest answers.

Is this just an AI summary?

No. Every claim is anchored to the source, the timestamp, and a screenshot, and the pipeline is built to avoid inventing numbers. The difference is "an AI summarized it" versus "here is the claim and the exact frame it came from." You can act on the second one.

How is this different from using ChatGPT on a transcript?

Transcripts miss the screen. Half the value in a build tutorial is on-screen: the file tree, the config, the dashboard, the exact prompt. We capture every frame, so the copy-paste config isn't lost. And we run it every week on your channels, governed and consistent, so nothing slips.

What channels can you cover?

Any public video channel. We're proven on the AI-builder and AI-automation niche specifically, so if you follow Claude Code, AI agents, or automation-agency creators, you're exactly who this was built for.

What's the free first brief?

Pick one channel. We distill its recent uploads and send you a real brief at no cost and no card. If it saves you an evening of watching, you'll know.

Get started

Pick one channel.
Get your first brief free.

No card. We detect the signal and send it back. If it saves you an evening, you'll keep going.

One channel, one free brief. No card, no spam.
We use your email only to send your brief. Pulse is an independent research tool, not affiliated with or endorsed by YouTube or any featured creator.