Oshen Davidson (@_oshdavidson) profile picture

independent researcher | evals & studying AIs impact on human behavior & vice versa

pdx, or
Joined August 2025
Lo peor y más desesperante de un apocalipsis tecnológico es que lo primero que va dejar de funcionar es internet y no vamos a poder hacer chistes y memes mientras vemos el mundo arder.
140
4,984
163
51,359
338,091
Oshen Davidson retweeted
A good reminder that just because you cuda doesn’t mean you shuda.
38
220
24
2,781
79,839
Oshen Davidson retweeted
Is this management? Am I a manager now??
148
135
26
2,980
66,533
Oshen Davidson retweeted
adding her to my tailnet to show that i'm serious about the relationship
37
291
22
3,423
76,359
Oshen Davidson retweeted
"AI is just matrix multiplication" is the new "emotions are just chemicals in the brain"
157
439
60
5,612
147,630
Oshen Davidson retweeted
human-in-the-loop └ human-on-the-loop └ human-near-the-loop └ human-cc'd-on-the-loop └ human-legally-liable-for-the-loop
124
722
93
8,950
262,636
Oshen Davidson retweeted
the new version of “turn it off and turn it on again” is swapping models and trying again when a task fails
8
4
4
65
5,383
Oshen Davidson retweeted
I love working on AI safety because I get to benefit the most important issue of our time, arguably. If I wasn't working on AI safety, I suspect I'd feel far more stressed and powerless about AI.
10
7
2
164
8,793
Oshen Davidson retweeted
Replying to @github
Me listening to repos on a walkman
18
57
7
2,585
117,161
Oshen Davidson retweeted
When we evaluated the specific queries in Gemini 3's fan outs, this data point stood out to me most... most of the fan out queries had NO MSV. Just adding one or two words to a fan out, takes it from "head keyword" to something else. seerinteractive.com/insights…
3
4
1
28
9,335
Oshen Davidson retweeted
Meet the new @AirOpsHQ. The only all-in-one content platform to rank your brand #1 on ChatGPT. We've raised a $40M Series B, led by @GreylockVC People skip Google and ask ChatGPT "what's the best [your category]?" And if your brand doesn't show up in the answer, you lose. Here's how brands like Webflow, Chime, and Ramp use AirOps to win: 1. AirOps analyzes how often ChatGPT, other AI search engines, and traditional search recommend your brand. Page360 shows how each page performs across SEO, AI search, and your website's Google Analytics. 2. We highlight exactly where competitors get recommended and you don’t, then give you a prioritized roadmap so you know what to fix or create next. 3. AirOps connects to tools like Zendesk, HubSpot, and Gong, turning your data into inputs your team can use as context to make new content. 4. Automated workflows scale high-quality content, with human review at 5+ points. 5. In one click, publish agent-readable content straight into your website. 8-person marketing teams that once published 40 content pieces a month now publish 500+ with AirOps and outrank everyone. This is how modern brands get found in the new AI era of discovery. If you don’t 3× your AI traffic in six months, we’ll refund you up to $1M. Book a demo here: airops.com/book-a-call Since you’re still reading, here’s a little bonus 👇🏻
204
95
21
394
162,864
Oshen Davidson retweeted
If AI content is will eventually becoming 99% of content, how do you stand out? Was talking with a founder about this recently and here’s how you win distribution in a world full of bots: 1. Proprietary data + data visualizations win I think about how companies like SimilarWeb or AirOps have interesting data to share around web analytics or SEO. That data is often pulled from their own products and distilled in a way that’s easy to understand and therefore easy to distribute. Guess what type of people eat this content up? Mid-market to enterprise companies. 2. Talk strategy > tactics Tactics are easy to spew out via AI, but when your content is geared towards forward thinking strategy, it’s original. I’ve found that when I talk strategy on conference stages, it compels people to think deeper and therefore reach out to us for help. The tactical talks get high ratings and are filled with dopamine hits, but don’t do nearly as good of a job at driving pipeline. 3. The storyteller is the most powerful person in the room Guess who said that? Steve Jobs. And it’s never been more true today. If you can layer on data + strategy + strong storytelling, you’ll have the infinity stones of today’s distribution. Bonus: if you can tell stories that seem personalized to each individual, you’ll have the fourth infinity stones of distribution. What did I miss?
3
1
8
906
🔥 This Week’s Biggest SEO & AI Search Stories from #SEOFOMO 👇 * ​Reddit sues Perplexity, SerpApi over scraping Google Search data * ​OpenAI Introduces the ChatGPT Atlas web browser * ​Google working on fixing Search Console performance report delay * ​Google: Links, Site Moves & Technical SEO Don't Fix Quality Issues * ​How to achieve early clicks and bigger wins with AI before Black Friday * ​ChatGPT, LLM referrals convert worse than Google Search: Study * ​How to tell if you were impacted by the August 2025 spam update (and what to do about it) * ​Ha and you all said backlinks were dead * ​How and Why Google Rewrites Your Hard-Earned Headlines * ​How OpenAI’s Product Feed Redefines Commerce Data * ​The 2025 Digital Agency Survey * ​The Best of Published Decks from the Latest BrightonSEO [October 2025] * Much more! Including events, jobs, tools... by SEOs like @NickLeRoy @Marie_Haynes Harry Clarkson-Bennett Paddy Moogan @randfish @iPullRank Oshen Davidson and many more! Check it out: seofomo.co/posts/your-top-se…
5
7
2
30
3,519
🚨 Third-Party Sources Drive 85% of Brand Discovery in AI Search 👇 @AirOpsHQ conducted a research identifying that : * Third-Party Sources Account for 85% of Brand Mentions in AI Search * Listicles Drive 9 in 10 Third-Party Mentions in AI Search * Product Pages and Homepages Drive 26% of First-Party Visibility in AI Search * 68% of Brand Mentions Are Often Unique to a Single AI Model * Brand Recognition Compounds When Owned and Earned Channels Align "When on-site and off-site channels work together, visibility doesn’t just add up - it amplifies. Owned content provides the depth AI systems draw from, while earned visibility signals validation and relevance across the broader ecosystem." Read more: airops.com/report/the-influe…
3
13
1
45
10,016
Oshen Davidson retweeted
OpenAI just sidelined your website and changed SEO forever. It's about to unleash a whole new industry vertical in search optimization. All CMOs need to ship v1 of this on a war-footing: [thread]
8
9
51
17,540
We’re introducing the Agentic Commerce Protocol, an open standard co-developed with @Stripe that enables programmatic commerce flows between users, AI agents, and businesses.
ChatGPT already helps millions of people find what to buy. Now it can help them buy it too. We’re introducing Instant Checkout in ChatGPT with @Etsy and @Shopify, and open-sourcing the Agentic Commerce Protocol that powers it, built with @Stripe, so more merchants and developers can integrate agentic checkout.
98
243
60
2,019
408,440
Oshen Davidson retweeted
Big news -> Microsoft is in talks with US publishers about launching a two-sided marketplace that would compensate publishers for their content used by AI products "Microsoft would become the first major tech company to build an AI marketplace for publishers, a milestone in building a sustainable business model for content companies in the AI era." "The new Publisher Content Marketplace (PCM), as Microsoft is calling it, will launch as a pilot program with a limited set of publishers, according to publishing executives briefed on the matter." axios.com/2025/09/23/microso…
5
17
6
77
11,334
Oshen Davidson retweeted
Please read the whole thing. It's just 3 points. 1. LLMs *do understand schema* when they see it in input, contrary to a popular belief that the structured data is lost due to tokenization. This is because they're trained on it as they are trained on code. 2. LLMs *don't* always see schema in input because most tools and functions supplying context to the model strip it away during pre-processing. (The next part is where SEO people get confused.) 3. The tools AI systems use to process content which is passed as context to LLMs sometimes parse the HTML page by using structured data to better understand the page and sometimes they don't. It's a software design choice and nothing to do with LLM's capability. TL;DR: LLMs understand structured data if they get it, but most of the time they get plain text / markdown, which *may* or *may not* have been rendered by leveraging schema at the pre-processing level.
48
16
214
22,763