К списку видео

Alexander Whedon

Introducing SubQ - a major breakthrough in LLM intelligence. It is t...

Запуск

Платформа: X
Длительность: 1:24
Язык в кадре (предположительно): английский
Просмотры: 13,5 млн
Медианная склейка: 1.14s
Лицо в кадре: 60%
Тёмные кадры: 78%

Full video footage is a Pro feature. The stills are free: the source film is $10/mo.

More from Alexander Whedon

Dark-Mode Explainer Interview

medium effort

Steal this
Pillarbox the interview so the black side bars are the exact same black as your full-frame graphic cards. The talking head and the animation then live in one continuous dark canvas, and you can cut between them every second without the frame ever flashing.

A single seated presenter interview is chopped into ~1-second slices and interleaved with full-frame, near-black motion graphics that visualise each claim. The presenter footage is pillarboxed (vertical crop centred in a 16:9 frame) so the black side bars match the black graphic cards, letting the edit cut between talking head and animation without any visual jolt. Burned-in two-or-three-word captions run along the bottom of both layers, tying the whole thing together.

Opening move

Cold open on a full-frame near-black graphic — a titled benchmark bar chart, thin serif/light sans headline top-left, bars rising into frame with one bar accent-green against neutral cream — before any face appears. The claim is shown as a chart first, so the viewer is reading a result rather than being introduced to a person, and the black background instantly signals 'product launch' rather than 'vlog'.

Pacing

Relentless: 17 detected shot changes across 85s, median 1.14s between cuts. Assume roughly one cut per sentence fragment, with several sub-second flashes inside the animated grid sequence (three of the sampled frames sit within one second of each other around the 38s mark, i.e. a hard cut or a fast animated push on nearly every beat). Alternate blocks — two or three quick presenter sizes, then a graphic slab, then back. 78% of frames read dark, so let the graphics dominate screen time (18 no-face shots vs 27 presenter shots) and keep the face on screen only about 60% of the time.

Graphics and type

Two-tone system: near-black background, cream/off-white as the neutral fill, one accent colour per idea (green for the hero bar, yellow highlight strokes on documents, a single saturated blue disc, one red square). Headlines in a light serif or thin sans, top-left, generously letter-spaced, with a small subhead beneath. Chart labels tiny and unbold — deliberately understated, even slightly overlapping. Captions are burned in, centred at the bottom, medium-weight white sans, no box or stroke, only two or three words per card so they change on almost every cut. Presenter footage is graded low-key and slightly desaturated: cool green background, warm skin, crushed shadows, white shirt kept just under clipping. Vignette or natural falloff darkens the frame edges so the pillarbox blends into the graphics.

Shot list

Full-frame graphic card on near-black (#0d0d0d), headline top-left, chart or icon field centred
Carries the numeric/abstract claims and covers audio edits in the interview
Medium presenter, seated, centred in a pillarboxed vertical crop, head-and-shoulders with chest visible, shallow depth, dark green shelving falling to bokeh behind
Home base — establishes a person and authority between graphic slabs
Close presenter, same setup punched in to a tighter head crop, eyes on upper third, slight off-centre framing
Punch-in on emphasis lines; lets you cut mid-sentence from one size to another
Close presenter with white node-graph animation composited into the negative space beside the head, small caption label under it
Merges the two layers so the graphics feel native to the interview, not intercut B-roll
Minimal graphic: single blue disc with thin concentric arcs and hairline radial ticks, dead centre
Scale/geometry beat — one idea, one shape, high contrast against black
Extreme-minimal graphic: two tiny red and grey squares centred in an almost empty black frame
Visual full stop / breathing beat before the next block, and the closing frame
Grid of cream document icons with yellow highlight strokes, slowly drifting/scrolling vertically
Volume metaphor — motion continues across three consecutive cuts so the tempo stays up without new shots

How to shoot it

Lock the camera on a seated position, subject 2m+ in front of a dark green or otherwise deep-toned shelving wall. Light with one soft key front-left, kill spill on the background, then add a single small backlight so the set doesn't go muddy.
Shoot the whole script in one continuous take at a medium size, then reshoot the same script (or key lines) punched in tighter. Two sizes are enough to sustain a 1-second cut rhythm.
Decide your aspect trick first: crop the presenter footage to a vertical or square window centred in a 16:9 timeline, leaving pure black side bars. Set the graphic cards to the exact same black so cuts don't flash.
Build a graphics kit of five or six reusable cards on that black: a titled bar chart, an icon grid, a circle-with-arcs diagram, a node graph on transparent background, and a near-empty two-square frame. Use cream as the only neutral and one accent colour per card.
Edit the audio first as a tight radio cut with no pauses, then place a cut on nearly every sentence fragment — target a ~1.1s median. Rotate medium, close, graphic, close, graphic so no two adjacent shots match.
Where a claim is numeric or abstract, replace the face entirely with a full-frame graphic; where it's about the product working, composite the node graph into the empty space beside the head instead of cutting away.
Animate the icon grid with slow continuous vertical drift so you can cut three times inside it and still feel forward motion.
Auto-caption, then manually break the subtitles to two or three words maximum, centred bottom, plain white, no background — so the text changes as often as the picture.
Grade: crush the blacks to match the graphic background, desaturate the greens slightly, keep skin warm, add a soft vignette. Every frame should read dark.
End on the most minimal graphic in the set — the two tiny squares on black — held as the final frame rather than a logo lockup.

Kit

Mirrorless camera with a fast prime (35mm or 50mm equivalent) shot wide open for background separation
Tripod, locked off — reframe by moving the camera between takes or crop in post for the punch-in
One large soft key (5-in-1 or 60cm+ softbox) placed front-left at head height, plus a weak fill or bounce on the shadow side
A practical or small hair light on the shelving behind to lift the green background out of black
Lavalier or shotgun mic on a boom just out of frame
Dressed background: dark painted shelving with sparse, low-contrast props (books, small objects) at least 2m behind the subject
Motion graphics software (After Effects, or Keynote/Figma plus a simple animator) and an auto-caption tool for the burned-in subtitles

Not knowable from stills

Any music bed, sound design on the graphic transitions, or whether the presenter is sync-sound versus a separately recorded voiceover laid over silent performance
Whether the graphics were animated in After Effects or generated/rendered by another tool
Whether the punch-in close-ups are a second camera, a second take, or a digital crop of the same take
Exact frame rate and whether the sub-second graphic beats are true cuts or animated pushes within one shot
Precise hex values and typeface names for the headline and caption fonts

Written by anthropic/claude-opus-5 from the frames and measurements above. Technique only, and it cannot hear the video.

Video MCPlive

Give your coding agent video taste.

A read-only MCP server over every indexed frame. No API key, no account.

Wire up your agent

Показано 35 из 45 · 10 свёрнуто в удержанные планы

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:00

medium presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:00

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:07

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:07

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:08

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:08

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:10

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:10

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:13

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:13

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:14

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:14

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:16

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:16

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:18

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:18

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:19

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:19

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:22

medium presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:22

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:28

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:28

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:31

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:31

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:34

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:34

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:36

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:36

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:37

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:37

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:37

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:37

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:37

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:37

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:37

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:37

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:37

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:37

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:38

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:38

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:40

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:40

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:40

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:40

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:40

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:40

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:41

medium presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:41

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:46

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:46

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:49

medium presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:49

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:55

medium presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:55

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 0:56

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

0:56

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 1:01

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

1:01

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 1:04

medium presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

1:04

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 1:07

medium presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

1:07

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 1:10

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

1:10

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 1:16

medium presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

1:16

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 1:22

close presenter

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

1:22

Alexander Whedon Introducing SubQ - a major breakthrough in LLM intelligence. It is t... at 1:22

Alexander WhedonIntroducing SubQ - a major breakthrough in LLM intelligence. It is t...

1:22

Расшифровка: Introducing SubQ - a major breakthrough in LLM intelligence. It is t...

Расшифровано речевой моделью по звуку самого ролика, а не написано брендом. Нажмите на строку, чтобы перейти к этому моменту.

0:00Today we're announcing a major architectural breakthrough, the world's first fully subquadratic
0:04LLM. SubQ has a 12 million token context window in frontier-level intelligence. It outperforms
0:10Opus 4.7 on long context at less than 5% of the cost. At 1 million tokens, SubQ processes tokens
0:1752 times faster than flashed attention and dramatically changes the cost of training and
0:21inference, enabling us to scale context windows to 12 million tokens and beyond, pushing the
0:26boundaries of intelligence. This means your agents can now complete weeks of work at a time without
0:30degrading, reasoning across entire codebases, merging hundreds of PRs at once, and finding
0:35patterns in tens of thousands of documents, all in one shot, without losing accuracy, speed, or context.
0:41SubQ is not just another model. It represents a major algorithmic breakthrough. It is the first
0:46model built on a fully subquadratic, sparse attention architecture. LLMs today waste compute
0:51by processing every possible relationship between words,
0:54but only a small fraction of these matter.
0:57SubQ finds and focuses only on those relationships,
1:00meaning compute is used where it matters most.
1:03At 12 million tokens, this reduces compute almost 1,000 times,
1:07changing the way the LLMs scale.
1:09Today, SubQ is available for early access,
1:11along with SubQ Code,
1:13a coding agent built on the next generation of AI technology.
1:16This marks a new era for builders,
1:17an era powered by subquadratic intelligence.
1:19Visit subq.ai to get early access.

22 строк · 238 слов

Cut at the same speed

Films whose median shot runs closest to 1.14s.

Remove the friction Magnific Desktop is live for macOS and Windows ...

Remove the friction Magnific Desktop is live for macOS and Windows ...

0:3932 frames1.1s cut44% face

november was a certified banger.

november was a certified banger.

1:0068 frames1.1s cut78% face

Build your entire business on Base44

Build your entire business on Base44

0:5254 frames1.1s cut26% face

France x Jacquemus

France x Jacquemus

0:4743 frames1.2s cut74% face

Hims Big Game Commercial: "Rich People Live Longer"

Hims Big Game Commercial: "Rich People Live Longer"

1:0058 frames1.2s cut55% face

MiniMax H3 is now available on Runway. Access it alongside every oth...

Launch69K views

MiniMax H3 is now available on Runway. Access it alongside every oth...

0:1110 frames1.2s cut10% face

Newsletters Are the Easiest Business You Can Start Right Now

Newsletters Are the Easiest Business You Can Start Right Now

123:07140 frames1.2s cut57% face

Pierre Niney & Pierre Gasly for Lacoste & Alpine.

Pierre Niney & Pierre Gasly for Lacoste & Alpine.

0:4133 frames1.2s cut33% face

He Built an AI Shopping App to 15,000 Users in 2 Months

He Built an AI Shopping App to 15,000 Users in 2 Months

1:5096 frames1.1s cut65% face

Lo importante está aquí

Lo importante está aquí

0:1514 frames1.1s cut86% face

Lift Up, Take Off | Peloton App

Lift Up, Take Off | Peloton App

0:1519 frames1.2s cut63% face

THE CREATIVE PROCESS, a film by Gawx

THE CREATIVE PROCESS, a film by Gawx

1:4282 frames1.2s cut18% face

Lit and framed like this

Closest on how much of the running time has a face in it and how dark the frames are.

How Bloom Built the Playbook Every Founder Will Copy in 2026

How Bloom Built the Playbook Every Founder Will Copy in 2026

127:40162 frames0.5s cut60% face

He Built a $725 Million Company in 255 Days. Here's What You're Doing Wrong

He Built a $725 Million Company in 255 Days. Here's What You're Doing Wrong

118:51155 frames0.6s cut58% face

Ever wish you could scribble a prompt instead of needing to write it?...

Ever wish you could scribble a prompt instead of needing to write it?...

0:4435 frames1.7s cut60% face

La Famiglia Eyewear

La Famiglia Eyewear

0:3430 frames1.6s cut63% face

How Oakberry Became the Fastest-Growing Franchise You’ve Never Studied

How Oakberry Became the Fastest-Growing Franchise You’ve Never Studied

113:05158 frames0.5s cut65% face

Fat Loss Scientist: It’s Easy To Lose Weight, But Here’s Why You WON’T Do It! | Dr Andy Galpin

The Diary Of A CEO

Fat Loss Scientist: It’s Easy To Lose Weight, But Here’s Why You WON’T Do It! | Dr Andy Galpin

132:19162 frames1.4s cut63% face

Scared Money Is Exactly Why You're Not Growing (ft. Michael Brandt)

Scared Money Is Exactly Why You're Not Growing (ft. Michael Brandt)

135:51169 frames0.8s cut65% face

Professor Brian Greene: The Threat of AI, Consciousness & The End of Humanity

The Diary Of A CEO

Professor Brian Greene: The Threat of AI, Consciousness & The End of Humanity

131:44156 frames2.0s cut56% face

9-Figure Founder: How To Build a Brand That's Impossible To Copy

70:09165 frames0.2s cut54% face

Pinterest | You Just Might Surprise Yourself

Pinterest | You Just Might Surprise Yourself

1:1739 frames2.3s cut67% face

That idea you keep meaning to build. @itskasturiii used Codex to fin...

OpenAI Developers

That idea you keep meaning to build. @itskasturiii used Codex to fin...

1:2350 frames2.9s cut56% face

How 3 Students Built a €130K ARR AI Learning System

How 3 Students Built a €130K ARR AI Learning System

2:22139 frames2.7s cut66% face

More launch

Same category, other makers, most watched first.

Introducing Base44’s Big Game Ad “It’s App to You”

Launch140M views

Introducing Base44’s Big Game Ad “It’s App to You”

0:3024 frames1.9s cut79% face

Introducing Base44’s Big Game Ad Teaser “It’s App to You”

Launch138M views

Introducing Base44’s Big Game Ad Teaser “It’s App to You”

0:1515 frames1.5s cut67% face

Introducing Perplexity Computer. Computer unifies every current AI c...

Launch38M views

Introducing Perplexity Computer. Computer unifies every current AI c...

1:3432 frames3% face

Today we're excited to introduce Devin, the first AI software enginee...

Launch31M views

Today we're excited to introduce Devin, the first AI software enginee...

1:4937 frames11% face

introducing friend. not imaginary. order now at

Launch27M views

introducing friend. not imaginary. order now at

1:4450 frames2.9s cut50% face

Introducing Sakana Fugu: A full multi-agent orchestration system acce...

Launch26M views

Introducing Sakana Fugu: A full multi-agent orchestration system acce...

5 frames0% face

Introducing Gemini Omni: Create Anything from Anything

Launch26M views

Introducing Gemini Omni: Create Anything from Anything

0:5462 frames0.8s cut8% face

It's no secret how we feel about typing at Wispr Flow. The truth is,...

Launch19M views

It's no secret how we feel about typing at Wispr Flow. The truth is,...

2:00110 frames5.1s cut52% face

Announcing a new division of Midjourney called "Midjourney Medical"

Launch19M views

Announcing a new division of Midjourney called "Midjourney Medical"

2:23199 frames5.6s cut0% face

Introducing AI Key, a small device that lets AI control your entire p...

Launch15M views

Introducing AI Key, a small device that lets AI control your entire p...

1:1936 frames6.6s cut53% face

Announcing Personal Computer. Personal Computer is an always on, loc...

Launch14M views

Announcing Personal Computer. Personal Computer is an always on, loc...

1:2834 frames1.7s cut0% face

Introducing ChatGPT Images 2.0 A state-of-the-art image model that c...

Launch13M views

Introducing ChatGPT Images 2.0 A state-of-the-art image model that c...

1:0062 frames0.8s cut0% face

Browse every frame in the library