AJ Ghergich

My new pet goblin! Love the spritesheet and hatch skill! Look how adorable he is when he fails...not that he ever would fail

Thx for the reply Brian. The use case im hoping for is something like Notion can host bundled SKILL.md skills (SKILL.md plus references/*.md plus scripts/*.py) end-to-end via MCP etc. The brainlabs CEO article (I'm sure you read) using Notion was great, but it's not really using SKILLS the right way, IMO, until it can do the full bundle. Otherwise, it's just a prompt library.

@SEO @RayFernando1337 It would need a translation layer. I'm tinkering on this which supports syncing skill script files to Notion pages as code blocks: github.com/brianlovin/not…

Notion is landing this at the perfect time with the right abstractions. I’m excited to focus on building cool experiences on top of this.

New post from yours truly. The Case for Cloaking for Large Language Models. ipullrank.com/cloaking-for-l… *Cue outrage.*

You've been asking for this one... Now in preview: Codex in the ChatGPT mobile app. Start new work, review outputs, steer execution, and approve next steps, all from the ChatGPT mobile app. Codex will keep running on your laptop, Mac mini, or devbox.

@brian_lovin @RayFernando1337 Can notion host and distribute SKILLS now with code not just markdown files

@RayFernando1337 lmk what you hack on 👀

I spent the last 3 weeks running what might be the most comprehensive LLM ranking factors analysis to date. 29,562 unique domains tracked and scored across 145 industries, 1,595 buyer personas, and 105k+ ChatGPT prompts. Over 500TB of data, and 12 external signals correlated against rank-weighted LLM recommendation scores. This is a large-scale correlation study: what external signals actually predict whether a brand gets recommended by ChatGPT, across 145 industries and 1,595 buyer personas. -- Research Process 145 industries from 500 candidates. 11 personas each (10 targeted + 1 neutral). 25 runs per persona, rank-weighted scoring. 29,562 unique domains tracked. Data collected: - Common Crawl: 1.15B pages, domain mentions + phrase co-occurrences - Reddit: 5B+ posts and comments scanned - Google Search: 15,697 queries, top 100 results; 1.5M+ results captured - SERP HTML: parsed for outbound links and phrase presence - Wikimedia: 300M+ Wikidata entities + Wikipedia citations - Backlinks (Common Crawl Web Graph): PageRank + Harmonic Centrality; 4B+ - Top Site Homepages: parsed for persona-specific phrases -- Analysis Process 13 signals per domain. Spearman ρ vs. LLM recommendation score, per-industry and globally. R² shows variance explained. Lift measures over-representation in the top 10% most-recommended domains. Tiered: Dominant (ρ ≥ 0.30) down to Baseline (< 0.05). -- Key Findings SERP appearances, SERP rank, and outbound links from search results pages are the three strongest signals. Traditional SEO is the dominant measurable influence on LLM recommendations. Backlink authority (PageRank, Harmonic Centrality) follows. Combined, these point to one thing: established search authority drives LLM visibility. Signal hierarchies vary by industry. Wikidata dominates in established categories (hotels, ERP, furniture). Reddit drives community-driven ones (enterprise AI, live entertainment). No universal strategy. 80–85% of recommendation variance is inside the model. All external signals combined explain under 20%. You cannot infer LLM visibility from search rankings; you have to test it directly. -- The Two Conclusions That Matter 1. SEO is the foundation. OpenAI is using search data today and building their own index. As that matures, the connection between search authority and LLM visibility deepens. Traditional SEO principles are not obsolete, they're the starting point for LLM visibility too. 2. Persona is the measurement unit. The #1 airline for a frequent flyer is a different site from the #1 for a student flying abroad. Same model, same industry, different person, different result. You don't have one LLM rank, you have a rank per buyer segment. Monitor by persona or the number is meaningless. -- Full Report and data for all 145 industries and 1,595 personas available here: oppalerts.com/LLM-Ranking-Fa…

FAQ schema died twice from spam. The second time was last Wednesday. The fix is a new schema .org type: FAQSection. I filed it. joost.blog/faq-schema-cyc…

Many SEOs might be interested in using Claude Code to build agents + automations. This video showcases how you can set up a basic one using Claude's skill files.

Google to no longer support FAQ rich results searchengineland.com/google-to-no-l…

@chongdashu @jaredatch +1 I want to build a game with my 10-year-old this summer, so a tutorial would be amazing

@jaredatch That's a great idea! Kids have great imaginations. It would be great to bring their game ideas to life.

Just a couple of prompts from start to finish into a playable game! Pixel-perfect art > Codex App w/ GPT 5.5 High > Images 2.0 for animations (except walkcycles) > WAN 2.0 for walkcycles (cheap!) > 11Labs for bgm/sfx (in 1 prompt) > Phaser 4 Sound ON🔉 Full tutorial anyone?

Me pretending I understand sub-quadratic sparse-attention architecture (SSA)

@robin_liquidium @ajambrosino I know, but there is a million things they need to do on the normie side of the table...they should put 70% of the effort there for now.

@SEO @ajambrosino you can switch between simple and developer mode in settings. simple is the mode for normies

what should we do in the next 3 months?

@orinthomas Bring bro back

@SEO Those that forget BonzaiBuddy are doomed to reinvent it ;-p

My new pet goblin! Love the spritesheet and hatch skill! Look how adorable he is when he fails...not that he ever would fail

@ajambrosino Not for noobs u didn't though, I like the new mode, it's a start for sure. Or am I missing something?

@SEO we did

@gregberryai I wanted that experience badly lol..I got it working ok after a ton of tweaking and display port 1.4. Glad it's working for I thought. Size wise it's amazing

@SEO Interesting. I've had exactly 0 issues. It's, by far, the best monitor I've ever had.

thinking about getting the apple studio display 27", is it worth it? right now i have the lg 34" UltraWide 5K (about 7 years old, held up very well)

@IAmIrvSol Hot Goblin Summer, let's go

@SEO Goblin mode all summer 👺👹

@gregberryai @MatthewBerman Google it, it's notorious, YT videos explain all the problems

@SEO @MatthewBerman Connected to a MacBook Pro

@koltregaskes @david_saint_ Didn't Elon say multiple times the current size for Grok is 4.3 etc 500 and he gave the roadmap and sizes for each model?

@david_saint_ Hard to say but with the estimate of a much larger model for the Gemini range, I'm wondering if that's something to do with the TPUs.

GPT-5.5 likely has around 1.5T parameters after a sanity check of the Incompressible Knowledge Probes paper that originally claimed 9.7T. GPT-5.5 = 1.5T Claude Opus 4.7 = 1.1T GPT-5 = 1.3T Grok-4.20 = 768B Gemini 3.1 Pro = 4.7T Benjamin and Lawrence reproduced the code and regression before identifying an undocumented floor on low scores and serious ambiguities in about 25 percent of the hard tier questions. The IKP dataset pulls heavily from obscure researcher records via DBLP and OpenAlex where a quarter of tough questions proved ambiguous or had incorrect gold answers giving the revised estimates real punch. lesswrong.com/posts/veFMEzDD…

@WyldeApes Let's go!

@SEO goblin/acc

@gregberryai @MatthewBerman I have this monitor...do not under any circumstances get this monitor for a Mac...ultrawides like this don't work well Apple

@MatthewBerman I highly recommend the Samsung 57" Odyssey Neo G9 (G95NC) Series Dual 4K UHD 1000R Curved Gaming Monitor. The thing is ridiculous in all the best ways. Super crisp, insane screen real estate.