Pinned Tweet
First SEO problem...
How to know what is page/keyword is ranking if you can't see them in GSC?
Peter Nikolow retweeted
JavaScript spent 15 years making websites worse.
Then invented an entire ecosystem of tools to make them work like normal websites again.
Imagine if we built houses the same way.
jonoalderson.com/misc/the-ho…
Peter Nikolow retweeted
BREAKING: President Trump calls on Jensen Huang to "make an apology and go back" to Nvidia, $NVDA, as CEO.
Trump also says he discussed the situation with China's President Xi.
Nvidia would reinstate Jensen as CEO "if they were smart," Trump says.
Peter Nikolow retweeted
5 new open-source agent skills this week for international/multilingual/technical SEO and AI search:
Bing grounding queries export + analysis, schema implementation audits, title/meta gap fill and international llms.txt builds.
All 5 and 13 more:
github.com/rebelytics/one-sk…
"Thousands of devices controlling tens of thousands of refrigeration devices were taken off the open internet either remotely or, in some cases, physically - with staff ordered to remove internet cables and monitor temperatures manually."
🧊
bbc.com/news/articles/c6eq49…
24
Peter Nikolow retweeted
Browser Run's /snapshot endpoint can now return four more formats from a single page load:
- PDFs
- Links
- Scrape
- JSON
developers.cloudflare.com/ch…
Peter Nikolow retweeted
BREAKING: we've officially entered a deal to acquire Mailchimp for $3.68 billion!
Peter Nikolow retweeted
Oh boy … PHP sent clear text data when you used uppercased S in httpS:// in a protocol handler 🙈🙈
file_get_contents('HTTPS://…') skipped TLS: tokens, POST bodies and FTPS logins went out unencrypted!
Fixed today in 8.2.36, 8.3.37, 8.4.28, 8.5.13 & 8.6.0.
AI mistakes will be expensive
Wall Street giants pledged to mobilize $500 billion to help Nvidia customers finance chip purchases. But no capital has been deployed to date.
"Jensen Huang sensed the winds shifting against Nvidia this spring...In August, he got pledges from several firms—Apollo, Blackstone, BlackRock, Brookfield, Goldman Sachs and KKR—to mobilize an astounding sum: $500 billion...To date, no capital has actually been deployed..."
wsj.com/finance/wall-street-… via @WSJ
1
103
SEO is fine. The future of AI is tied to search.
Read: dejan.ai/blog/seo-is-fine/
An LLM on its own isn't all knowing. They store enough knowledge in their parametric memory to 'get by' and act intelligently, but the dynamic and real-time memory lives in search indexes. Any claim outside of that is pure science fiction.
Video Credit: youtube.com/watch?v=ctcMA6ch…
When you're deciding which version of Gemini to use, keep this in mind:
Hydrogen (H) — Atomic number 1
Carbon (C) — Atomic number 6
Neon (Ne) — Atomic number 10 <--- released today
Argon (Ar) — Atomic number 18
If you want to know how your brand is doing in Neon, you can use this:
dejan.ai/tools/treewalker/ru…
Algoroo V2, combines traditional and AI search flux: dejan.ai/algoroo/
Peter Nikolow retweeted
If you want 768GB of HBM3 VRAM at a reasonable price, now is your chance to lock it in. Otherwise you'll get caught paying way more for a DGX Station with "unified memory" tinycorp.myshopify.com/produ…
Peter Nikolow retweeted
Tracking the Nginx HTTP/2 CVE packages here
So far ... we're all still screwed 😂
Disable HTTP/2 in the meantime!
ma.ttias.be/nginx-http2-cve-…
Peter Nikolow retweeted
PSA: be VERY CAREFUL what third party JS you're deploying on production.
Where possible, always control the source.
Peter Nikolow retweeted
In November 1995, Vermeer Technologies developed the FrontPage 1.0 editor for WYSIWYG web development.
In January 1996, Microsoft bought Vermeer and, in June, released an editor called Microsoft FrontPage 1.1
Peter Nikolow retweeted
New CXL 2.0 card lets DDR5 servers reuse eight DDR4 memory modules
🔗videocardz.com/newz/new-cxl-…
Peter Nikolow retweeted
Introducing Qwen3.8-27B-pi-adapted 🤗
We adapted Qwen's model specifically for the pi coding agent using reinforcement learning and a LoRA adapter.
The results, measured as pass@1:
🚀 SWE-bench Pro (Hard 51): 29.4% → 62.7%
💻 Terminal-Bench 2.1: 39.5% → 55.8%
⚡ 55–62% fewer tool calls per task
This is the power of open models: you can adapt them to YOUR tools and YOUR use cases. Let’s build! 💛
Open weights: huggingface.co/harness-hub/Q…
Methodology: huggingface.co/blog/harness-…





























