@ai.with.andrew69k followers25d agotalking-head

This open source model from Google lets you predict anything in the future.

9.0×its baseline

40k
views: 4.9k
baseline: 2.1×
engagement: 2.1k
likes: 2.3k
saves: 524
shares

Why it beat the baseline

hook · transcript
“This open source model from Google lets you predict anything in the future.”

talking-headutility

Bold claim about a Google open-source AI forecasting tool
Explains TimesFM mechanics, dataset size, and use cases
Contrasts legacy custom training costs with zero-shot forecasting
Highlights new confidence interval features and native spreadsheet integrations

Why it worked. Framed a complex time-series foundational model as an accessible superpower integrated directly into everyday spreadsheet workflows.

Grounded in the actual video and audio; analyzed by gemini-3.8-flash, cached.

Video analysis

Every shot, the spoken script, on-screen text, audio, editing and a recreation blueprint, from the actual video and audio.

talking-headshowcasehook: bold claim56s · 23 shots · 9:16speech · music · captionscta: none

Hook (spoken)This open source model from Google lets you predict anything in the future.

Hook (on screen)ai predicts THE FUTURE

Why it stops the scrollStirs immediate curiosity for data professionals and general tech audiences wanting predictive foresight.

TechniqueCombines authority ('Google', 'open source') with an audacious promise ('predict anything in the future').

Formattalking head with fast technical b-roll cutaways and screen graphics

Settinghome office / studio with modern warm interior lighting

Subjectmale AI creator seated in front of a laptop and podcast microphone

Lookblack crewneck t-shirt, short dark hair, well-groomed beard

Lighting / gradewarm key light from side, soft fill; clean, natural with high-contrast UI graphics

Pacing / tonefast, punchy educational pace; informative, authoritative, engaging

AudioSingle male narrator, clean studio microphone quality, forward mix.. Music: Million Dolla Hip Hop by Spaceman; upbeat lo-fi hip hop beat with steady drums.. SFX: Subtle swooshes on slide transitions and document flips..

Editing22 cuts, 2.4s avg; Hard cuts mixed with downward quick slides on graphic elements.; captions: Synced lowercase bold white subtitles with intermittent large kinetic black titles.

Topicsartificial intelligence, machine learning, google ai, data science, productivity

Products / appsGoogle TimesFM (00:00.0: Core subject featured via GitHub repo and research diagrams.); Google Sheets (00:49.8: Logo graphic shown as supported platform.); Google BigQuery (00:50.8: Logo graphic shown as supported database.); RGB Studio Microphone (00:00.0: Desktop mic on small tripod in front of speaker.)

Structure · 56s

00:00.0–00:03.9The Hook. Grab attention with the claim of predicting the future via open-source AI.
00:03.9–00:16.8Product & Scope Introduction. Introduce TimesFM and ground it in relatable everyday forecasting examples.
00:16.8–00:37.7The Friction vs. Solution. Contrast time/cost of custom models with zero-training foundational inference.
00:37.7–00:48.5New Feature: Confidence Bands. Explain range predictions rather than static single-point numbers.
00:48.5–00:55.7Workplace Integration & Payoff. Showcase immediate availability in Sheets and BigQuery for non-technical teams.

Shot list

#	time	shot	what is visible	spoken	on screen
1	00:00.0–00:01.8	medium close-up split tripod; static	GitHub repo google-research/timesfm card above presenter speaking into mic Screen: TimesFM GitHub banner	This open source model from Google	ai predicts THE FUTURE · this open source · from Google
2	00:01.8–00:03.9	b-roll motion graphic digital render; dynamic zoom / particle drift	matrix-like number lines and black blocks shifting Screen: animated abstract data matrix	lets you predict anything in the future.	lets you predict · anything in the
3	00:03.9–00:05.7	screen record screen capture; smooth scroll down	TimesFM GitHub code repository and documentation Screen: github.com/google-research/timesfm README	This is Times FM,	this is Times · FM
4	00:05.7–00:07.7	motion graphic digital animation; rotating neon arc	Google AI glowing circular UI with 'Introducing AI Mode' Screen: Google Gemini circular animation	Google's new AI model for	Google's new AI · model for
5	00:07.7–00:09.1	graphic chart static graphic; static	Cumulative Return with Cost line charts Screen: evaluation charts from paper	time series forecasting	FORECASTING
6	00:09.1–00:11.8	graphic document motion graphics / screen scroll; pan down	text excerpt highlighting 100 billion real world time-points and transformer architecture Screen: TimesFM architecture paper snippet	trained on 100 billion real world data points.	ON · 100 · REAL
7	00:11.8–00:13.5	b-roll stock footage close up; subtle push	crypto and stock trading dashboard on tablet screen Screen: BTC candlestick trading graph	Think stock prices,	think stock prices
8	00:13.5–00:14.9	b-roll stock footage medium shot; static	dim bedroom, person waking up resting head on arm Screen: none	your sleep score each night,	your sleep score each night
9	00:14.9–00:16.8	b-roll stock footage medium shot; gentle pan	corporate team discussing sales numbers Screen: none	or even your business and sales forecast.	or even your · business and sales
10	00:16.8–00:21.8	motion graphics digital scroll; continuous downward scroll	spreadsheets, charts, data summaries scrolling quickly Screen: data science technical paper pages	Until now, you'd have to train a separate model with your own data every single time,	UNTIL · HAVE · DATA · SINGLE
11	00:21.8–00:24.4	b-roll stock footage medium rear angle; static	developer typing on dual monitor workstation Screen: IDE code windows	which takes weeks and costs a heck ton of money.	which takes · weeks and costs · of money
12	00:24.4–00:26.7	motion graphic digital render; subtle pulse	Google multicolor 'G' logo Screen: Google logo	But Times FM flips this completely.	but Times FM · flips this completely
13	00:26.7–00:28.9	medium close-up tripod; static	presenter explaining with open palm gestures Screen: none	You just give it your historical numbers	you just give · it your historical
14	00:28.9–00:31.9	graphic diagram digital pan; pan across diagram	token residual blocks and time feature layers Screen: TimesFM model architecture	and it forecasts the future with zero training required,	IT · FUTURE · ZERO
15	00:31.9–00:35.8	b-roll montage digital animation; rapid wipe transitions between planetary, bubble chamber, and celestial maps	crater terrain map, particle physics bubble tracks, vintage star chart Screen: scientific data visualizations	because it already learned patterns from billions of real data points	because it already · Learned patterns · from billions of · real data points
16	00:35.8–00:37.7	b-roll stock footage medium shot; lateral pan	modern office environments with office workers analyzing computer screen Screen: IDE code & business reports	across sales, traffic, energy and more.	sales · traffic · energy and more
17	00:37.7–00:40.7	b-roll / quote overlay pan; panning	social quote 'Math just made a little more sense. Thanks Claude.' over math formulas Screen: social media post graphic	And the new version doesn't just tell you what will happen,	and the new · version doesn't just · tell you what
18	00:40.7–00:43.0	b-roll interview clip medium shot; static	Sundar Pichai talking and gesturing Screen: none	it also tells you how confident it is.	it also tells · you how confident
19	00:43.0–00:48.5	motion graphics collage top-down stop motion graphic; stop-motion pop-in animation	tiled financial sheets and charts filling the screen Screen: stop-motion animation of spreadsheet reports	So instead of a single number, you get a full range from best case to worst case and everything in between.	so instead of · a single number · full range from · best case to · worst case · between
20	00:48.5–00:49.8	medium close-up tripod; static	creator gesturing with palms up Screen: none	It's already live	it's already live
21	00:49.8–00:51.6	graphics animation motion graphics; pop-in cut	Google Sheets icon followed by BigQuery icon Screen: Google Sheets & BigQuery logos	inside Google Sheets and BigQuery,	Google Sheets · BigQuery
22	00:51.6–00:53.7	split screen tripod & stock footage; static	three suited executives looking attentively at a screen above creator Screen: finance chart on computer screen	which means your finance team can forecast next quarter	your finance · forecast next
23	00:53.7–00:55.7	split screen tripod & stock footage; static	programmer typing code above presenter talking to camera Screen: developer screen code	without writing a single line of code.	without writing a · single line of

Script

00:00.0–00:01.8“This open source model from Google” · emphatic and high energy
00:01.8–00:03.9“lets you predict anything in the future.” · dramatic emphasis
00:03.9–00:07.7“This is Times FM, Google's new AI model for” · informative and clear
00:07.7–00:11.8“time series forecasting trained on 100 billion real world data points.” · rhythmic cadence emphasizing numbers
00:11.8–00:14.0“Think stock prices,” · conversational example
00:14.0–00:15.0“your sleep score each night,” · conversational example
00:15.0–00:16.8“or even your business and sales forecast.” · grounded corporate application
00:16.8–00:21.8“Until now, you'd have to train a separate model with your own data every single time,” · contrasting, problem-focused cadence
00:21.8–00:24.4“which takes weeks and costs a heck ton of money.” · relatable frustration
00:24.4–00:26.7“But Times FM flips this completely.” · punchy turning point
00:26.7–00:28.9“You just give it your historical numbers” · casual, easy description
00:28.9–00:31.9“and it forecasts the future with zero training required,” · impressive claim emphasis
00:31.9–00:35.8“because it already learned patterns from billions of real data points” · explaining rationale clearly
00:35.8–00:37.7“across sales, traffic, energy and more.” · rapid listing
00:37.7–00:40.7“And the new version doesn't just tell you what will happen,” · building anticipation
00:40.7–00:43.0“it also tells you how confident it is.” · highlighting key feature
00:43.0–00:48.5“So instead of a single number, you get a full range from best case to worst case and everything in between.” · practical explanation
00:48.5–00:51.6“It's already live inside Google Sheets and BigQuery,” · direct actionable news
00:51.6–00:55.7“which means your finance team can forecast next quarter without writing a single line of code.” · conclusive value proposition

On-screen text

00:00.0–00:02.1“ai predicts THE FUTURE” · hook · center · bold black sans-serif hook headline
00:00.0–00:01.2“this open source” · caption · mid-lower third · white bold sans-serif burned subtitles
00:01.2–00:01.8“from Google” · caption · mid-lower third · white bold sans-serif burned subtitles
00:01.8–00:02.7“lets you predict” · caption · mid-lower third · white bold sans-serif burned subtitles
00:02.7–00:03.9“anything in the” · caption · mid-lower third · white bold sans-serif burned subtitles
00:03.9–00:04.9“this is Times” · caption · mid-lower third · white bold sans-serif burned subtitles
00:04.9–00:05.7“FM” · label · bottom center · bold large black uppercase text
00:05.7–00:06.8“Google's new AI” · caption · mid-lower third · white bold sans-serif burned subtitles
00:06.8–00:07.7“model for” · caption · mid-lower third · white bold sans-serif burned subtitles
00:07.7–00:09.1“FORECASTING” · caption · bottom center · large black bold uppercase word-by-word graphic
00:09.1–00:09.7“ON” · caption · bottom center · large black bold uppercase word-by-word graphic
00:09.7–00:10.8“100” · caption · bottom center · large black bold uppercase word-by-word graphic
00:10.8–00:11.8“REAL” · caption · bottom center · large black bold uppercase word-by-word graphic
00:11.8–00:13.5“think stock prices” · caption · lower third · white bold sans-serif burned subtitles
00:13.5–00:14.9“your sleep score each night” · caption · lower third · white bold sans-serif burned subtitles
00:14.9–00:15.8“or even your” · caption · lower third · white bold sans-serif burned subtitles
00:15.8–00:16.8“business and sales” · caption · lower third · white bold sans-serif burned subtitles
00:16.8–00:18.0“UNTIL” · caption · bottom center · large black bold uppercase word-by-word graphic
00:18.0–00:19.4“HAVE” · caption · bottom center · large black bold uppercase word-by-word graphic
00:19.4–00:20.6“DATA” · caption · bottom center · large black bold uppercase word-by-word graphic
00:20.6–00:21.8“SINGLE” · caption · bottom center · large black bold uppercase word-by-word graphic
00:21.8–00:22.5“which takes” · caption · lower third · white bold sans-serif burned subtitles
00:22.5–00:23.5“weeks and costs” · caption · lower third · white bold sans-serif burned subtitles
00:23.5–00:24.4“of money” · caption · lower third · white bold sans-serif burned subtitles
00:24.4–00:25.6“but Times FM” · caption · lower third · white bold sans-serif burned subtitles
00:25.6–00:26.7“flips this completely” · caption · lower third · white bold sans-serif burned subtitles
00:26.7–00:27.7“you just give” · caption · lower third · white bold sans-serif burned subtitles
00:27.7–00:28.9“it your historical” · caption · lower third · white bold sans-serif burned subtitles
00:28.9–00:29.8“IT” · caption · bottom center · large black bold uppercase word graphic
00:29.8–00:30.8“FUTURE” · caption · bottom center · large black bold uppercase word graphic
00:30.8–00:31.9“ZERO” · caption · bottom center · large black bold uppercase word graphic
00:31.9–00:32.8“because it already” · caption · lower third · white bold sans-serif burned subtitles
00:32.8–00:33.8“Learned patterns” · caption · lower third · white bold sans-serif burned subtitles
00:33.8–00:34.9“from billions of” · caption · lower third · white bold sans-serif burned subtitles
00:34.9–00:35.8“real data points” · caption · lower third · white bold sans-serif burned subtitles
00:35.8–00:36.5“sales” · caption · lower third · white bold sans-serif burned subtitles
00:36.5–00:37.1“traffic” · caption · lower third · white bold sans-serif burned subtitles
00:37.1–00:37.7“energy and more” · caption · lower third · white bold sans-serif burned subtitles
00:37.7–00:38.6“and the new” · caption · lower third · white bold sans-serif burned subtitles
00:38.6–00:39.9“version doesn't just” · caption · lower third · white bold sans-serif burned subtitles
00:39.9–00:40.7“tell you what” · caption · lower third · white bold sans-serif burned subtitles
00:40.7–00:41.8“it also tells” · caption · lower third · white bold sans-serif burned subtitles
00:41.8–00:43.0“you how confident” · caption · lower third · white bold sans-serif burned subtitles
00:43.0–00:43.9“so instead of” · caption · lower third · white bold sans-serif burned subtitles
00:43.9–00:44.9“a single number” · caption · lower third · white bold sans-serif burned subtitles
00:44.9–00:45.9“full range from” · caption · lower third · white bold sans-serif burned subtitles
00:45.9–00:46.8“best case to” · caption · lower third · white bold sans-serif burned subtitles
00:46.8–00:47.7“worst case” · caption · lower third · white bold sans-serif burned subtitles
00:47.7–00:48.5“between” · caption · lower third · white bold sans-serif burned subtitles
00:48.5–00:49.8“it's already live” · caption · lower third · white bold sans-serif burned subtitles
00:49.8–00:50.8“Google Sheets” · label · bottom center below logo · bold black sans-serif app label
00:50.8–00:51.6“BigQuery” · label · bottom center below logo · bold black sans-serif app label
00:51.6–00:52.5“your finance” · caption · lower third · white bold sans-serif burned subtitles
00:52.5–00:53.7“forecast next” · caption · lower third · white bold sans-serif burned subtitles
00:53.7–00:54.6“without writing a” · caption · lower third · white bold sans-serif burned subtitles
00:54.6–00:55.7“single line of” · caption · lower third · white bold sans-serif burned subtitles

Recreation blueprint

55s AI release breakdown combining talking-head commentary, b-roll use-cases, and paper diagrams.

slot	s	visual	voiceover	overlay
Hook	3.9	Split talking head with GitHub repo banner, cutting to abstract data animation.	Bold statement claiming a tech giant model allows anyone to forecast anything.	ai predicts THE FUTURE
Model Intro & Everyday Scenarios	12.9	Repo screen record, architecture snippet, and 3 quick b-roll real-world examples.	Name the model, cite the massive training scale, and list 3 practical business use cases.	FORECASTING / 100 BILLION
Old Way Pain vs New Way	20.9	Paper scrolls, late-night coder, transitioning back to presenter and schematic.	Contrast costly custom training with zero-shot instant forecasting on raw history.	UNTIL NOW / ZERO TRAINING
Differentiator & Distribution	18	Stop-motion spreadsheet report montage followed by app logos and split-screen.	Detail confidence interval ranges and reveal built-in integration into everyday tools.	Google Sheets / BigQuery

Keep

Rapid b-roll cutaways every 1.5–2.5 seconds
High-contrast text overlays highlighting scale and numbers
Direct juxtaposition between legacy model training pain and zero-shot ease

Can change

Specific AI model covered (e.g., vision, LLM, speech)
Target enterprise applications and tools mentioned

Generation prompts per shot (3)

Shot 1. Vertical format medium close-up of a young male creator speaking directly to camera with glowing RGB desk microphone, laptop open, studio lighting, top half showing software repo card, modern tech aesthetic.
Negative: blurry, low resolution, noisy, bad lighting
Shot 2. Abstract digital data visualization, rows of flowing numbers and monochrome matrix blocks against vintage paper texture background, clean graphic render.
Negative: messy, 3d characters, low res
Shot 19. Stop-motion top-down view of crumpled office paper, printed financial graphs, and bright green circular stickers assembling neatly on a black tabletop, modern graphic design style.
Negative: photorealistic clutter, dark, illegible

Raw JSON (for a video pipeline)

{
  "hook": {
    "spoken": "This open source model from Google lets you predict anything in the future.",
    "ends_at": "00:03.9",
    "on_screen": "ai predicts THE FUTURE",
    "starts_at": "00:00.0",
    "technique": "bold_claim",
    "technique_note": "Combines authority ('Google', 'open source') with an audacious promise ('predict anything in the future').",
    "why_it_stops_the_scroll": "Stirs immediate curiosity for data professionals and general tech audiences wanting predictive foresight."
  },
  "tags": {
    "topics": [
      "artificial intelligence",
      "machine learning",
      "google ai",
      "data science",
      "productivity"
    ],
    "cta_type": "none",
    "language": "en",
    "has_music": true,
    "has_speech": true,
    "content_type": "showcase",
    "products_or_apps": [
      "TimesFM",
      "Google Sheets",
      "Google BigQuery",
      "GitHub"
    ],
    "has_burned_captions": true
  },
  "audio": {
    "music": "Million Dolla Hip Hop by Spaceman; upbeat lo-fi hip hop beat with steady drums.",
    "voice": "Single male narrator, clean studio microphone quality, forward mix.",
    "ambient": "Silent studio room tone.",
    "mix_notes": "Voice dominates center stage; background instrumental stays ducked at around -14dB.",
    "sound_effects": "Subtle swooshes on slide transitions and document flips."
  },
  "shots": [
    {
      "end": "00:01.8",
      "index": 1,
      "start": "00:00.0",
      "camera": "tripod",
      "framing": "split screen with GitHub repo card on top and creator at bottom",
      "movement": "static",
      "shot_type": "medium close-up split",
      "transition_in": "none",
      "screen_content": "TimesFM GitHub banner",
      "subject_action": "gesturing with hands while talking",
      "what_is_visible": "GitHub repo google-research/timesfm card above presenter speaking into mic",
      "duration_seconds": 1.8,
      "spoken_during_shot": "This open source model from Google",
      "on_screen_text_during_shot": [
        "ai predicts THE FUTURE",
        "this open source",
        "from Google"
      ],
      "facial_expression_or_gesture": "expressive, engaged"
    },
    {
      "end": "00:03.9",
      "index": 2,
      "start": "00:01.8",
      "camera": "digital render",
      "framing": "full screen abstract animation",
      "movement": "dynamic zoom / particle drift",
      "shot_type": "b-roll motion graphic",
      "transition_in": "cut",
      "screen_content": "animated abstract data matrix",
      "subject_action": "none",
      "what_is_visible": "matrix-like number lines and black blocks shifting",
      "duration_seconds": 2.1,
      "spoken_during_shot": "lets you predict anything in the future.",
      "on_screen_text_during_shot": [
        "lets you predict",
        "anything in the"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:05.7",
      "index": 3,
      "start": "00:03.9",
      "camera": "screen capture",
      "framing": "GitHub repository page",
      "movement": "smooth scroll down",
      "shot_type": "screen record",
      "transition_in": "cut",
      "screen_content": "github.com/google-research/timesfm README",
      "subject_action": "none",
      "what_is_visible": "TimesFM GitHub code repository and documentation",
      "duration_seconds": 1.8,
      "spoken_during_shot": "This is Times FM,",
      "on_screen_text_during_shot": [
        "this is Times",
        "FM"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:07.7",
      "index": 4,
      "start": "00:05.7",
      "camera": "digital animation",
      "framing": "centered logo animation",
      "movement": "rotating neon arc",
      "shot_type": "motion graphic",
      "transition_in": "cut",
      "screen_content": "Google Gemini circular animation",
      "subject_action": "none",
      "what_is_visible": "Google AI glowing circular UI with 'Introducing AI Mode'",
      "duration_seconds": 2,
      "spoken_during_shot": "Google's new AI model for",
      "on_screen_text_during_shot": [
        "Google's new AI",
        "model for"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:09.1",
      "index": 5,
      "start": "00:07.7",
      "camera": "static graphic",
      "framing": "two time-series line graphs stacked",
      "movement": "static",
      "shot_type": "graphic chart",
      "transition_in": "cut",
      "screen_content": "evaluation charts from paper",
      "subject_action": "none",
      "what_is_visible": "Cumulative Return with Cost line charts",
      "duration_seconds": 1.4,
      "spoken_during_shot": "time series forecasting",
      "on_screen_text_during_shot": [
        "FORECASTING"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:11.8",
      "index": 6,
      "start": "00:09.1",
      "camera": "motion graphics / screen scroll",
      "framing": "academic paper excerpt and architecture diagram",
      "movement": "pan down",
      "shot_type": "graphic document",
      "transition_in": "cut",
      "screen_content": "TimesFM architecture paper snippet",
      "subject_action": "none",
      "what_is_visible": "text excerpt highlighting 100 billion real world time-points and transformer architecture",
      "duration_seconds": 2.7,
      "spoken_during_shot": "trained on 100 billion real world data points.",
      "on_screen_text_during_shot": [
        "ON",
        "100",
        "REAL"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:13.5",
      "index": 7,
      "start": "00:11.8",
      "camera": "close up",
      "framing": "hands using stylus on financial tablet with candlestick chart",
      "movement": "subtle push",
      "shot_type": "b-roll stock footage",
      "transition_in": "cut",
      "screen_content": "BTC candlestick trading graph",
      "subject_action": "drawing projection trendline on chart",
      "what_is_visible": "crypto and stock trading dashboard on tablet screen",
      "duration_seconds": 1.7,
      "spoken_during_shot": "Think stock prices,",
      "on_screen_text_during_shot": [
        "think stock prices"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:14.9",
      "index": 8,
      "start": "00:13.5",
      "camera": "medium shot",
      "framing": "person lying in bed under duvet holding head",
      "movement": "static",
      "shot_type": "b-roll stock footage",
      "transition_in": "cut",
      "screen_content": "none",
      "subject_action": "waking up",
      "what_is_visible": "dim bedroom, person waking up resting head on arm",
      "duration_seconds": 1.4,
      "spoken_during_shot": "your sleep score each night,",
      "on_screen_text_during_shot": [
        "your sleep score each night"
      ],
      "facial_expression_or_gesture": "tired"
    },
    {
      "end": "00:16.8",
      "index": 9,
      "start": "00:14.9",
      "camera": "medium shot",
      "framing": "three business colleagues in conference room around laptop",
      "movement": "gentle pan",
      "shot_type": "b-roll stock footage",
      "transition_in": "cut",
      "screen_content": "none",
      "subject_action": "pointing at screen in discussion",
      "what_is_visible": "corporate team discussing sales numbers",
      "duration_seconds": 1.9,
      "spoken_during_shot": "or even your business and sales forecast.",
      "on_screen_text_during_shot": [
        "or even your",
        "business and sales"
      ],
      "facial_expression_or_gesture": "serious, focused"
    },
    {
      "end": "00:21.8",
      "index": 10,
      "start": "00:16.8",
      "camera": "digital scroll",
      "framing": "rapid flip through complex research PDF tables and output files",
      "movement": "continuous downward scroll",
      "shot_type": "motion graphics",
      "transition_in": "cut",
      "screen_content": "data science technical paper pages",
      "subject_action": "none",
      "what_is_visible": "spreadsheets, charts, data summaries scrolling quickly",
      "duration_seconds": 5,
      "spoken_during_shot": "Until now, you'd have to train a separate model with your own data every single time,",
      "on_screen_text_during_shot": [
        "UNTIL",
        "HAVE",
        "DATA",
        "SINGLE"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:24.4",
      "index": 11,
      "start": "00:21.8",
      "camera": "medium rear angle",
      "framing": "developer working in dark office on code editors",
      "movement": "static",
      "shot_type": "b-roll stock footage",
      "transition_in": "cut",
      "screen_content": "IDE code windows",
      "subject_action": "coding late at night",
      "what_is_visible": "developer typing on dual monitor workstation",
      "duration_seconds": 2.6,
      "spoken_during_shot": "which takes weeks and costs a heck ton of money.",
      "on_screen_text_during_shot": [
        "which takes",
        "weeks and costs",
        "of money"
      ],
      "facial_expression_or_gesture": "concentrated"
    },
    {
      "end": "00:26.7",
      "index": 12,
      "start": "00:24.4",
      "camera": "digital render",
      "framing": "centered Google G logo on black background",
      "movement": "subtle pulse",
      "shot_type": "motion graphic",
      "transition_in": "cut",
      "screen_content": "Google logo",
      "subject_action": "none",
      "what_is_visible": "Google multicolor 'G' logo",
      "duration_seconds": 2.3,
      "spoken_during_shot": "But Times FM flips this completely.",
      "on_screen_text_during_shot": [
        "but Times FM",
        "flips this completely"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:28.9",
      "index": 13,
      "start": "00:26.7",
      "camera": "tripod",
      "framing": "creator centered behind desk and mic",
      "movement": "static",
      "shot_type": "medium close-up",
      "transition_in": "cut",
      "screen_content": "none",
      "subject_action": "talking enthusiastically",
      "what_is_visible": "presenter explaining with open palm gestures",
      "duration_seconds": 2.2,
      "spoken_during_shot": "You just give it your historical numbers",
      "on_screen_text_during_shot": [
        "you just give",
        "it your historical"
      ],
      "facial_expression_or_gesture": "enthusiastic"
    },
    {
      "end": "00:31.9",
      "index": 14,
      "start": "00:28.9",
      "camera": "digital pan",
      "framing": "architecture schematic with transformer causal self-attention",
      "movement": "pan across diagram",
      "shot_type": "graphic diagram",
      "transition_in": "cut",
      "screen_content": "TimesFM model architecture",
      "subject_action": "none",
      "what_is_visible": "token residual blocks and time feature layers",
      "duration_seconds": 3,
      "spoken_during_shot": "and it forecasts the future with zero training required,",
      "on_screen_text_during_shot": [
        "IT",
        "FUTURE",
        "ZERO"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:35.8",
      "index": 15,
      "start": "00:31.9",
      "camera": "digital animation",
      "framing": "circular viewport vignette showing scientific maps and astronomy charts",
      "movement": "rapid wipe transitions between planetary, bubble chamber, and celestial maps",
      "shot_type": "b-roll montage",
      "transition_in": "cut",
      "screen_content": "scientific data visualizations",
      "subject_action": "none",
      "what_is_visible": "crater terrain map, particle physics bubble tracks, vintage star chart",
      "duration_seconds": 3.9,
      "spoken_during_shot": "because it already learned patterns from billions of real data points",
      "on_screen_text_during_shot": [
        "because it already",
        "Learned patterns",
        "from billions of",
        "real data points"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:37.7",
      "index": 16,
      "start": "00:35.8",
      "camera": "medium shot",
      "framing": "two developers discussing code at monitor, then two colleagues analyzing paper at desk",
      "movement": "lateral pan",
      "shot_type": "b-roll stock footage",
      "transition_in": "cut",
      "screen_content": "IDE code & business reports",
      "subject_action": "working and reviewing results",
      "what_is_visible": "modern office environments with office workers analyzing computer screen",
      "duration_seconds": 1.9,
      "spoken_during_shot": "across sales, traffic, energy and more.",
      "on_screen_text_during_shot": [
        "sales",
        "traffic",
        "energy and more"
      ],
      "facial_expression_or_gesture": "engaged"
    },
    {
      "end": "00:40.7",
      "index": 17,
      "start": "00:37.7",
      "camera": "pan",
      "framing": "person walking in front of architectural grid wall, then chalkboard equation background with quote pop-up",
      "movement": "panning",
      "shot_type": "b-roll / quote overlay",
      "transition_in": "cut",
      "screen_content": "social media post graphic",
      "subject_action": "walking past background",
      "what_is_visible": "social quote 'Math just made a little more sense. Thanks Claude.' over math formulas",
      "duration_seconds": 3,
      "spoken_during_shot": "And the new version doesn't just tell you what will happen,",
      "on_screen_text_during_shot": [
        "and the new",
        "version doesn't just",
        "tell you what"
      ],
      "facial_expression_or_gesture": "thoughtful"
    },
    {
      "end": "00:43.0",
      "index": 18,
      "start": "00:40.7",
      "camera": "medium shot",
      "framing": "Sundar Pichai seated in interview on grey chair",
      "movement": "static",
      "shot_type": "b-roll interview clip",
      "transition_in": "cut",
      "screen_content": "none",
      "subject_action": "speaking at conference",
      "what_is_visible": "Sundar Pichai talking and gesturing",
      "duration_seconds": 2.3,
      "spoken_during_shot": "it also tells you how confident it is.",
      "on_screen_text_during_shot": [
        "it also tells",
        "you how confident"
      ],
      "facial_expression_or_gesture": "confident"
    },
    {
      "end": "00:48.5",
      "index": 19,
      "start": "00:43.0",
      "camera": "top-down stop motion graphic",
      "framing": "crumpled papers, data sheets, bar charts and green stickers multiplying",
      "movement": "stop-motion pop-in animation",
      "shot_type": "motion graphics collage",
      "transition_in": "cut",
      "screen_content": "stop-motion animation of spreadsheet reports",
      "subject_action": "none",
      "what_is_visible": "tiled financial sheets and charts filling the screen",
      "duration_seconds": 5.5,
      "spoken_during_shot": "So instead of a single number, you get a full range from best case to worst case and everything in between.",
      "on_screen_text_during_shot": [
        "so instead of",
        "a single number",
        "full range from",
        "best case to",
        "worst case",
        "between"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:49.8",
      "index": 20,
      "start": "00:48.5",
      "camera": "tripod",
      "framing": "creator talking to camera",
      "movement": "static",
      "shot_type": "medium close-up",
      "transition_in": "cut",
      "screen_content": "none",
      "subject_action": "speaking",
      "what_is_visible": "creator gesturing with palms up",
      "duration_seconds": 1.3,
      "spoken_during_shot": "It's already live",
      "on_screen_text_during_shot": [
        "it's already live"
      ],
      "facial_expression_or_gesture": "excited"
    },
    {
      "end": "00:51.6",
      "index": 21,
      "start": "00:49.8",
      "camera": "motion graphics",
      "framing": "centered logo sequence on white background",
      "movement": "pop-in cut",
      "shot_type": "graphics animation",
      "transition_in": "cut",
      "screen_content": "Google Sheets & BigQuery logos",
      "subject_action": "none",
      "what_is_visible": "Google Sheets icon followed by BigQuery icon",
      "duration_seconds": 1.8,
      "spoken_during_shot": "inside Google Sheets and BigQuery,",
      "on_screen_text_during_shot": [
        "Google Sheets",
        "BigQuery"
      ],
      "facial_expression_or_gesture": "none"
    },
    {
      "end": "00:53.7",
      "index": 22,
      "start": "00:51.6",
      "camera": "tripod & stock footage",
      "framing": "top half shows executives peering at screen; bottom half shows creator speaking",
      "movement": "static",
      "shot_type": "split screen",
      "transition_in": "cut",
      "screen_content": "finance chart on computer screen",
      "subject_action": "creator gesturing, executives looking closely",
      "what_is_visible": "three suited executives looking attentively at a screen above creator",
      "duration_seconds": 2.1,
      "spoken_during_shot": "which means your finance team can forecast next quarter",
      "on_screen_text_during_shot": [
        "your finance",
        "forecast next"
      ],
      "facial_expression_or_gesture": "intent, expressive"
    },
    {
      "end": "00:55.7",
      "index": 23,
      "start": "00:53.7",
      "camera": "tripod & stock footage",
      "framing": "top half shows engineer coding; bottom half shows creator wrapping up",
      "movement": "static",
      "shot_type": "split screen",
      "transition_in": "cut",
      "screen_content": "developer screen code",
      "subject_action": "creator speaking concluding thought",
      "what_is_visible": "programmer typing code above presenter talking to camera",
      "duration_seconds": 2,
      "spoken_during_shot": "without writing a single line of code.",
      "on_screen_text_during_shot": [
        "without writing a",
        "single line of"
      ],
      "facial_expression_or_gesture": "smiling / persuasive"
    }
  ],
  "video": {
    "pacing": "fast, punchy educational pace",
    "setting": "home office / studio with modern warm interior lighting",
    "subject": "male AI creator seated in front of a laptop and podcast microphone",
    "lighting": "warm key light from side, soft fill",
    "color_grade": "clean, natural with high-contrast UI graphics",
    "orientation": "vertical",
    "aspect_ratio": "9:16",
    "overall_format": "talking head with fast technical b-roll cutaways and screen graphics",
    "tone_and_energy": "informative, authoritative, engaging",
    "duration_seconds": 55.7,
    "wardrobe_and_look": "black crewneck t-shirt, short dark hair, well-groomed beard"
  },
  "script": [
    {
      "end": "00:01.8",
      "line": "This open source model from Google",
      "start": "00:00.0",
      "delivery": "emphatic and high energy",
      "uncertain": false
    },
    {
      "end": "00:03.9",
      "line": "lets you predict anything in the future.",
      "start": "00:01.8",
      "delivery": "dramatic emphasis",
      "uncertain": false
    },
    {
      "end": "00:07.7",
      "line": "This is Times FM, Google's new AI model for",
      "start": "00:03.9",
      "delivery": "informative and clear",
      "uncertain": false
    },
    {
      "end": "00:11.8",
      "line": "time series forecasting trained on 100 billion real world data points.",
      "start": "00:07.7",
      "delivery": "rhythmic cadence emphasizing numbers",
      "uncertain": false
    },
    {
      "end": "00:14.0",
      "line": "Think stock prices,",
      "start": "00:11.8",
      "delivery": "conversational example",
      "uncertain": false
    },
    {
      "end": "00:15.0",
      "line": "your sleep score each night,",
      "start": "00:14.0",
      "delivery": "conversational example",
      "uncertain": false
    },
    {
      "end": "00:16.8",
      "line": "or even your business and sales forecast.",
      "start": "00:15.0",
      "delivery": "grounded corporate application",
      "uncertain": false
    },
    {
      "end": "00:21.8",
      "line": "Until now, you'd have to train a separate model with your own data every single time,",
      "start": "00:16.8",
      "delivery": "contrasting, problem-focused cadence",
      "uncertain": false
    },
    {
      "end": "00:24.4",
      "line": "which takes weeks and costs a heck ton of money.",
      "start": "00:21.8",
      "delivery": "relatable frustration",
      "uncertain": false
    },
    {
      "end": "00:26.7",
      "line": "But Times FM flips this completely.",
      "start": "00:24.4",
      "delivery": "punchy turning point",
      "uncertain": false
    },
    {
      "end": "00:28.9",
      "line": "You just give it your historical numbers",
      "start": "00:26.7",
      "delivery": "casual, easy description",
      "uncertain": false
    },
    {
      "end": "00:31.9",
      "line": "and it forecasts the future with zero training required,",
      "start": "00:28.9",
      "delivery": "impressive claim emphasis",
      "uncertain": false
    },
    {
      "end": "00:35.8",
      "line": "because it already learned patterns from billions of real data points",
      "start": "00:31.9",
      "delivery": "explaining rationale clearly",
      "uncertain": false
    },
    {
      "end": "00:37.7",
      "line": "across sales, traffic, energy and more.",
      "start": "00:35.8",
      "delivery": "rapid listing",
      "uncertain": false
    },
    {
      "end": "00:40.7",
      "line": "And the new version doesn't just tell you what will happen,",
      "start": "00:37.7",
      "delivery": "building anticipation",
      "uncertain": false
    },
    {
      "end": "00:43.0",
      "line": "it also tells you how confident it is.",
      "start": "00:40.7",
      "delivery": "highlighting key feature",
      "uncertain": false
    },
    {
      "end": "00:48.5",
      "line": "So instead of a single number, you get a full range from best case to worst case and everything in between.",
      "start": "00:43.0",
      "delivery": "practical explanation",
      "uncertain": false
    },
    {
      "end": "00:51.6",
      "line": "It's already live inside Google Sheets and BigQuery,",
      "start": "00:48.5",
      "delivery": "direct actionable news",
      "uncertain": false
    },
    {
      "end": "00:55.7",
      "line": "which means your finance team can forecast next quarter without writing a single line of code.",
      "start": "00:51.6",
      "delivery": "conclusive value proposition",
      "uncertain": false
    }
  ],
  "editing": {
    "cut_count": 22,
    "transitions": "Hard cuts mixed with downward quick slides on graphic elements.",
    "caption_style": "Synced lowercase bold white subtitles with intermittent large kinetic black titles.",
    "end_card_or_cta": "None; ends abruptly on punchline.",
    "average_shot_seconds": 2.4,
    "effects_and_overlays": "Split screens, circular vignettes, motion graphics overlays, stop-motion paper collage."
  },
  "teardown": {
    "hook": "This open source model from Google lets you predict anything in the future.",
    "format": "talking-head",
    "mechanism": "utility",
    "structure": [
      "Bold claim about a Google open-source AI forecasting tool",
      "Explains TimesFM mechanics, dataset size, and use cases",
      "Contrasts legacy custom training costs with zero-shot forecasting",
      "Highlights new confidence interval features and native spreadsheet integrations"
    ],
    "hook_source": "transcript",
    "why_it_worked": "Framed a complex time-series foundational model as an accessible superpower integrated directly into everyday spreadsheet workflows."
  },
  "on_screen_text": [
    {
      "end": "00:02.1",
      "kind": "hook",
      "text": "ai predicts THE FUTURE",
      "start": "00:00.0",
      "style": "bold black sans-serif hook headline",
      "source": "creator",
      "position": "center"
    },
    {
      "end": "00:01.2",
      "kind": "caption",
      "text": "this open source",
      "start": "00:00.0",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "mid-lower third"
    },
    {
      "end": "00:01.8",
      "kind": "caption",
      "text": "from Google",
      "start": "00:01.2",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "mid-lower third"
    },
    {
      "end": "00:02.7",
      "kind": "caption",
      "text": "lets you predict",
      "start": "00:01.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "mid-lower third"
    },
    {
      "end": "00:03.9",
      "kind": "caption",
      "text": "anything in the",
      "start": "00:02.7",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "mid-lower third"
    },
    {
      "end": "00:04.9",
      "kind": "caption",
      "text": "this is Times",
      "start": "00:03.9",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "mid-lower third"
    },
    {
      "end": "00:05.7",
      "kind": "label",
      "text": "FM",
      "start": "00:04.9",
      "style": "bold large black uppercase text",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:06.8",
      "kind": "caption",
      "text": "Google's new AI",
      "start": "00:05.7",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "mid-lower third"
    },
    {
      "end": "00:07.7",
      "kind": "caption",
      "text": "model for",
      "start": "00:06.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "mid-lower third"
    },
    {
      "end": "00:09.1",
      "kind": "caption",
      "text": "FORECASTING",
      "start": "00:07.7",
      "style": "large black bold uppercase word-by-word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:09.7",
      "kind": "caption",
      "text": "ON",
      "start": "00:09.1",
      "style": "large black bold uppercase word-by-word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:10.8",
      "kind": "caption",
      "text": "100",
      "start": "00:09.7",
      "style": "large black bold uppercase word-by-word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:11.8",
      "kind": "caption",
      "text": "REAL",
      "start": "00:10.8",
      "style": "large black bold uppercase word-by-word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:13.5",
      "kind": "caption",
      "text": "think stock prices",
      "start": "00:11.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:14.9",
      "kind": "caption",
      "text": "your sleep score each night",
      "start": "00:13.5",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:15.8",
      "kind": "caption",
      "text": "or even your",
      "start": "00:14.9",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:16.8",
      "kind": "caption",
      "text": "business and sales",
      "start": "00:15.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:18.0",
      "kind": "caption",
      "text": "UNTIL",
      "start": "00:16.8",
      "style": "large black bold uppercase word-by-word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:19.4",
      "kind": "caption",
      "text": "HAVE",
      "start": "00:18.0",
      "style": "large black bold uppercase word-by-word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:20.6",
      "kind": "caption",
      "text": "DATA",
      "start": "00:19.4",
      "style": "large black bold uppercase word-by-word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:21.8",
      "kind": "caption",
      "text": "SINGLE",
      "start": "00:20.6",
      "style": "large black bold uppercase word-by-word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:22.5",
      "kind": "caption",
      "text": "which takes",
      "start": "00:21.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:23.5",
      "kind": "caption",
      "text": "weeks and costs",
      "start": "00:22.5",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:24.4",
      "kind": "caption",
      "text": "of money",
      "start": "00:23.5",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:25.6",
      "kind": "caption",
      "text": "but Times FM",
      "start": "00:24.4",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:26.7",
      "kind": "caption",
      "text": "flips this completely",
      "start": "00:25.6",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:27.7",
      "kind": "caption",
      "text": "you just give",
      "start": "00:26.7",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:28.9",
      "kind": "caption",
      "text": "it your historical",
      "start": "00:27.7",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:29.8",
      "kind": "caption",
      "text": "IT",
      "start": "00:28.9",
      "style": "large black bold uppercase word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:30.8",
      "kind": "caption",
      "text": "FUTURE",
      "start": "00:29.8",
      "style": "large black bold uppercase word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:31.9",
      "kind": "caption",
      "text": "ZERO",
      "start": "00:30.8",
      "style": "large black bold uppercase word graphic",
      "source": "creator",
      "position": "bottom center"
    },
    {
      "end": "00:32.8",
      "kind": "caption",
      "text": "because it already",
      "start": "00:31.9",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:33.8",
      "kind": "caption",
      "text": "Learned patterns",
      "start": "00:32.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:34.9",
      "kind": "caption",
      "text": "from billions of",
      "start": "00:33.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:35.8",
      "kind": "caption",
      "text": "real data points",
      "start": "00:34.9",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:36.5",
      "kind": "caption",
      "text": "sales",
      "start": "00:35.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:37.1",
      "kind": "caption",
      "text": "traffic",
      "start": "00:36.5",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:37.7",
      "kind": "caption",
      "text": "energy and more",
      "start": "00:37.1",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:38.6",
      "kind": "caption",
      "text": "and the new",
      "start": "00:37.7",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:39.9",
      "kind": "caption",
      "text": "version doesn't just",
      "start": "00:38.6",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:40.7",
      "kind": "caption",
      "text": "tell you what",
      "start": "00:39.9",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:41.8",
      "kind": "caption",
      "text": "it also tells",
      "start": "00:40.7",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:43.0",
      "kind": "caption",
      "text": "you how confident",
      "start": "00:41.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:43.9",
      "kind": "caption",
      "text": "so instead of",
      "start": "00:43.0",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:44.9",
      "kind": "caption",
      "text": "a single number",
      "start": "00:43.9",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:45.9",
      "kind": "caption",
      "text": "full range from",
      "start": "00:44.9",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:46.8",
      "kind": "caption",
      "text": "best case to",
      "start": "00:45.9",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:47.7",
      "kind": "caption",
      "text": "worst case",
      "start": "00:46.8",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:48.5",
      "kind": "caption",
      "text": "between",
      "start": "00:47.7",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:49.8",
      "kind": "caption",
      "text": "it's already live",
      "start": "00:48.5",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:50.8",
      "kind": "label",
      "text": "Google Sheets",
      "start": "00:49.8",
      "style": "bold black sans-serif app label",
      "source": "creator",
      "position": "bottom center below logo"
    },
    {
      "end": "00:51.6",
      "kind": "label",
      "text": "BigQuery",
      "start": "00:50.8",
      "style": "bold black sans-serif app label",
      "source": "creator",
      "position": "bottom center below logo"
    },
    {
      "end": "00:52.5",
      "kind": "caption",
      "text": "your finance",
      "start": "00:51.6",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:53.7",
      "kind": "caption",
      "text": "forecast next",
      "start": "00:52.5",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:54.6",
      "kind": "caption",
      "text": "without writing a",
      "start": "00:53.7",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    },
    {
      "end": "00:55.7",
      "kind": "caption",
      "text": "single line of",
      "start": "00:54.6",
      "style": "white bold sans-serif burned subtitles",
      "source": "creator",
      "position": "lower third"
    }
  ],
  "structure_beats": [
    {
      "end": "00:03.9",
      "beat": "The Hook",
      "start": "00:00.0",
      "purpose": "Grab attention with the claim of predicting the future via open-source AI."
    },
    {
      "end": "00:16.8",
      "beat": "Product & Scope Introduction",
      "start": "00:03.9",
      "purpose": "Introduce TimesFM and ground it in relatable everyday forecasting examples."
    },
    {
      "end": "00:37.7",
      "beat": "The Friction vs. Solution",
      "start": "00:16.8",
      "purpose": "Contrast time/cost of custom models with zero-training foundational inference."
    },
    {
      "end": "00:48.5",
      "beat": "New Feature: Confidence Bands",
      "start": "00:37.7",
      "purpose": "Explain range predictions rather than static single-point numbers."
    },
    {
      "end": "00:55.7",
      "beat": "Workplace Integration & Payoff",
      "start": "00:48.5",
      "purpose": "Showcase immediate availability in Sheets and BigQuery for non-technical teams."
    }
  ],
  "recreation_blueprint": {
    "shot_template": [
      {
        "slot": "Hook",
        "visual": "Split talking head with GitHub repo banner, cutting to abstract data animation.",
        "seconds": 3.9,
        "voiceover": "Bold statement claiming a tech giant model allows anyone to forecast anything.",
        "text_overlay": "ai predicts THE FUTURE"
      },
      {
        "slot": "Model Intro & Everyday Scenarios",
        "visual": "Repo screen record, architecture snippet, and 3 quick b-roll real-world examples.",
        "seconds": 12.9,
        "voiceover": "Name the model, cite the massive training scale, and list 3 practical business use cases.",
        "text_overlay": "FORECASTING / 100 BILLION"
      },
      {
        "slot": "Old Way Pain vs New Way",
        "visual": "Paper scrolls, late-night coder, transitioning back to presenter and schematic.",
        "seconds": 20.9,
        "voiceover": "Contrast costly custom training with zero-shot instant forecasting on raw history.",
        "text_overlay": "UNTIL NOW / ZERO TRAINING"
      },
      {
        "slot": "Differentiator & Distribution",
        "visual": "Stop-motion spreadsheet report montage followed by app logos and split-screen.",
        "seconds": 18,
        "voiceover": "Detail confidence interval ranges and reveal built-in integration into everyday tools.",
        "text_overlay": "Google Sheets / BigQuery"
      }
    ],
    "what_can_change": [
      "Specific AI model covered (e.g., vision, LLM, speech)",
      "Target enterprise applications and tools mentioned"
    ],
    "template_summary": "55s AI release breakdown combining talking-head commentary, b-roll use-cases, and paper diagrams.",
    "what_must_be_kept": [
      "Rapid b-roll cutaways every 1.5–2.5 seconds",
      "High-contrast text overlays highlighting scale and numbers",
      "Direct juxtaposition between legacy model training pain and zero-shot ease"
    ],
    "generation_prompts": [
      {
        "prompt": "Vertical format medium close-up of a young male creator speaking directly to camera with glowing RGB desk microphone, laptop open, studio lighting, top half showing software repo card, modern tech aesthetic.",
        "shot_index": 1,
        "negative_prompt": "blurry, low resolution, noisy, bad lighting"
      },
      {
        "prompt": "Abstract digital data visualization, rows of flowing numbers and monochrome matrix blocks against vintage paper texture background, clean graphic render.",
        "shot_index": 2,
        "negative_prompt": "messy, 3d characters, low res"
      },
      {
        "prompt": "Stop-motion top-down view of crumpled office paper, printed financial graphs, and bright green circular stickers assembling neatly on a black tabletop, modern graphic design style.",
        "shot_index": 19,
        "negative_prompt": "photorealistic clutter, dark, illegible"
      }
    ]
  },
  "props_apps_and_products": [
    {
      "name": "Google TimesFM",
      "how_used": "Core subject featured via GitHub repo and research diagrams.",
      "first_seen": "00:00.0"
    },
    {
      "name": "Google Sheets",
      "how_used": "Logo graphic shown as supported platform.",
      "first_seen": "00:49.8"
    },
    {
      "name": "Google BigQuery",
      "how_used": "Logo graphic shown as supported database.",
      "first_seen": "00:50.8"
    },
    {
      "name": "RGB Studio Microphone",
      "how_used": "Desktop mic on small tripod in front of speaker.",
      "first_seen": "00:00.0"
    }
  ]
}

model gemini-3.8-flash · analyzed from the actual video and audio · 2026-10-14 · timestamps are estimates

Caption

Google’s Times FM predicts the future with zero training by using 100 billion real-world data points. It delivers confidence ranges for forecasts in finance, sales, and more—now built into Google Sheets and BigQuery. #AI #AITools #TechNews #Forecasting

Is this your account? Ask us to remove @ai.with.andrew: we stop tracking it and delete what we hold about it within a month.

Get this for your own niche

Copy AI founders for free, scan it on your plan, and get teardowns and adaptations for every outlier.

Start with this niche

More outliers in AI founders

Full report

200×1.0M vs 4.9k base
@ai.with.andrew69k followers14d ago
Claude is offering you a free subscription for one year.
tutorialutilityvideo analyzed · 48seng 1.2× · 38k saves · 13k shares
Video analysis Open on TikTok
48×267k vs 5.6k base
@20vc_tok280k followers4d ago
FRANCE WANTS TO TAX YOUR INTERNET USAGE
talking-headrelatable painvideo analyzed · 75seng 2.0× · 1.4k saves · 2.8k shares
Video analysis Open on TikTok
17×236k vs 5.0k base
@bengusberg82k followers11d ago
The rumors are true. Google can now tell you used AI in your writing with 93% accuracy,
talking-headrelatable painvideo analyzed · 76seng 1.0× · 12k saves · 2.3k shares
Video analysis Open on TikTok
16×228k vs 5.0k base
@bengusberg82k followers8d ago
USE CLAUDE TO GET A JOB
talking-headutilityvideo analyzed · 60seng 1.2× · 14k saves · 3.2k shares
Video analysis Open on TikTok
14×60k vs 4.2k base
@evisdrenova52k followers10d ago
#greenscreen #ai #anthropic #claude #tech
videoeng 3.2× · 644 saves · 256 shares
Teardown Open on TikTok