JUNKYARD KING ⚔️ EP1 - "Training Day"
Here’s what I learned making the latest episode of my 80s-inspired action-adventure fantasy buddy comedy about a kid, a glowing sword, and a mechanical crow.
A thread 👇
In the last 6 months at @Ahrefs, we analyzed over 1 billion data points across 14 studies. Here's what we learned about AI search optimization:
1) "Best X" blog listicles are the single most prominent content format cited by AI chatbots. They make up 43.8% of all page types cited by ChatGPT specifically.
2) 67% of ChatGPT's top 1,000 citations come from sources marketers can't influence: Wikipedia (29.7%), homepages (23.8%), app stores (6.6%). Only 32.3% are influenceable content like educational pages, reviews, news, and blog posts.
3) 28.3% of ChatGPT's most-cited pages have zero Google organic visibility. These pages get cited repeatedly by ChatGPT despite not ranking in Google at all. A completely separate discovery layer.
4) ChatGPT only cites about 50% of the URLs it retrieves. It fetches dozens of pages per query but uses half as background context without attribution. This means that being retrieved and being cited are very different things.
5) Adding schema markup had zero meaningful impact on AI citations. AI Overviews actually dipped −4.6%, while AI Mode (+2.4%) and ChatGPT (+2.2%) showed changes indistinguishable from zero.
6) YouTube mentions have the highest correlation (0.737) with AI brand visibility out of all the factors we studied (including all the conventional SEO metrics like backlinks, page count, DR, etc). This held true for both Google-owned and OpenAI products.
7) AI Overviews reduce clicks to the #1 result by 58%. That’s up from 34.5% just 10 months earlier. The trend is accelerating.
8) 99.9% of AI Overviews appear on informational intent queries. Transactional, navigational, and local searches are almost entirely AIO-free. Shopping triggers AIOs just 3.2% of the time.
9) For a given search query, Google’s AI Mode and AI Overviews reach the same conclusions 86% of the time — but cite almost entirely different sources (only 13.7% citation overlap).
10) AI Overviews change every 2.15 days on average, with 70% of content differing between consecutive observations. But semantic similarity stays at 0.95. The words, sources, and entities constantly shuffle, but the actual meaning barely moves.
https://t.co/WBvKwMHPDX
This is a great set of GEO experiments by Thomas.
At @invideoOfficial - we ran most of thee and a few other- all in all over 15 experiments in the last 6 months.
Found success with some of them, while a few turned out to be massive time sink.
Our AI citations, & AIO is ahead of our peers and as we sharpen our strategy to get coverage for specific topics, our lead will only increase.
Anyone who still wants to know more about which experiments to start and other ones that aren’t mentioned here- feel free to DM.
Our AI short film, THE INTERVIEW, officially went to Cannes.
A film by Vinu Joseph and Harsh Tiwari.
It follows a man in a government program that pays people to interview for jobs they'll never get.
Bombay! We're hiring a Community Lead.
Someone who watches a rough draft from a 19-year-old and knows exactly which two frames to keep. Someone who's been in the WhatsApp groups, the late-night edits, the "is this any good?" DMs. The kind of person friends send their first edits to.
Someone who treats a budding filmmaker's first short like it matters. Because it does.
If you love the craft and you love the people learning it, this one's yours.
Bombay-based.
DM @raichucopper with your work, IG and LinkedIn.
If you know that person, send this their way.
@luisnomad You are absolutely right @luisnomad - with multi-shot features on SOTA models & Agents that remember context across the projects individual frames no longer matter. https://t.co/KrYr48gtsN
Our work has been accepted at ICML 2026 main track!
https://t.co/Q9HyvQB0R9
If you are interested in alignment and specifically realignment then do check out our work on resource constraint editing of LLM values!
@murari_ai@debdeeplikesai
we’re moving past prompting
this is about directing now
made this using Agent One by @invideoOfficial
a creative agent that lets you build full video workflows solo
from concept to final cut, all in one flow
less friction, more control, lower cost
With Agent One by invideo you get to be the director you’ve always wanted to be.
You give it context once and it remembers it forever across every generation.
Just like that.
No re-explaining the same lighting condition 500 times.
No re-feeding it the same character sheet a millions times.
It does the heavy-lifting so you can make real progress.
Honest review: As I mentioned, with the launch of Seedance 2.0, there will be lots of platforms offering Unlimited, but there are tradeoffs with this kind of offer. That's why I only back a few platforms. @invideoOfficial is one of them: no crazy hype marketing, just a rock-solid subscription with 99. 99% successful generations and fast inference. Kudos to @_sankyy and invideo team for maintaining transparency with users.
Unitree Spring Festival Gala Robots —a Full Release of Additional Details 🥳
Dozens of G1 robots achieved the world’s first fully autonomous humanoid robot cluster Kung Fu performance (with quick movement), pushing motion limits and setting multiple world firsts! H2 made striking appearances at both the Beijing main venue and the Yiwu sub-venue, clad in the Monkey King’s heavy armor and riding a “somersault cloud” played by B2W quadruped robot dogs, delivering New Year blessings from the clouds.