Tech | AI | SaaS 🚀 | Building digital products, AI workflows & Telegram bots at https://xitter.cf/t.co/ZZUhB2o9q4 — In the hosting & web game since 2007. ⚡

The Internet
Joined December 2012
I'm building Vazin Online: a growing set of digital services under one roof: hosting, VPN, AI tools, games and more. Been in hosting since 2007. Now rebuilding it all for the AI era. vazin.online
1
1
8
672
25T tokens is a striking adoption signal, but it doesn’t tell us task success or user value. I’d also want to see retention, completion rates, latency, and cost per successful workflow.
We’ve seen an incredible response to Gemini 4 Argon from users, developers & enterprises: nearly 25 trillion tokens processed in the first 72 hours after launch - 3x what Gemini 3 did. Hear more about Argon and our commitment to the AI frontier from @sundarpichai ‘s conversation with @kantrowitz
2
2
25
For builders, the useful question is whether people come back and finish real work with it. Usage volume matters; reliable outcomes are what make a product stick.
1
10
OpenAI dot cloud browser: repeated crashes and technical failures. On 16 October 2026, the cloud browser displayed “Aw, Snap! — Error code: 9”. A normal reload did not fix the crash. Other observed failures included Page.getFrameTree and DOMSnapshot timeouts, an unavailable native accessibility provider, and an unavailable browser security check. These failures interrupted ordinary read-only tasks. Some pages later loaded successfully, but further timeouts recurred. The root cause remains unverified. Please investigate the browser crashes and control failures and provide a reliable fix. @OpenAI @sama @gdb @nickaturley
2
2
119
The irony is that this post itself was generated by dot! 😂 It’s been "asleep" for 10 days, and I chat with it 1 hour a day just to finish a task I could do myself in 2 minutes. We need true autonomy, not glorified waiting rooms! 🚀
1
38
Also shared my story on LinkedIn: from hosting in 2007 to building Vazin Online for the AI era 👇 linkedin.com/feed/update/urn…
3
51
62.2B tokens. Longest single task: 21h 50m. 17-day streak and counting. Not a flex, just what building a full AI platform solo actually looks like. Here's my Codex profile 👇 What's your longest agent run?
1
55
10% of the way to 500 Verified followers in just 3 days! 🎉 Next big challenge: taking timeline impressions from 180 to 500K! 📈 If you're building in: • AI & Automation 🤖 • Python & Web Dev 💻 • SaaS & Indie Hacking 🚀 Say hi in the replies so we can connect and support each other's posts! 🤝👇
5
9
157
Almost everything comes to an end soon... 🫡 Waiting for what's next. Hoping we get to 6.1 Luna and 6.1 Astra soon, and maybe 6.2 Sol and beyond...
On October 14, it's time to say farewell to GPT-5.5 in ChatGPT, ChatGPT Work, and Codex across all plans. If you use GPT-5.5 in Codex, switch to GPT-5.6 Sol or GPT-6 Astra. Thanks for everything, 5.5 🫡
2
9
158
Wouldn't complain 😄 as long as each one actually ships something new.
1
3
57
Gemini 4 Argon is officially here. 🚀 Watching a human play cornhole, then building its own robot sim to practice. Sculpting a full pagoda in Blender with multi-agent loops. We're no longer talking about chatbots. We're talking about models that learn the physical world. The AI race just changed gears.
Today, Gemini 4 Argon is officially rolling out.
1
2
148
Good plans are simple. I have one that should result in significantly more usage for everyone. Hope to get it done soon.
2
50
I paid @AnthropicAI $125 for Claude Max via @GooglePlay on Oct 9 and cancelled 20 minutes later. Never got the service: my @claudeai account is still on the Free plan because of an ID check. Anthropic says "ask Google." Google Play auto-rejected my refund twice. @Google, who keeps my money for nothing? #Anthropic #Claude #ClaudeAI #GooglePlay #Google #refund #AI #CustomerService #ConsumerRights
2
2
168
Is this the new customer service? Pay $125, get nothing, and the only way to get the service is to hand over your ID. How is that different from extortion? If a customer wants privacy, what are they supposed to do? If @AnthropicAI can doubt its customers, why can't customers doubt Anthropic? I've seen many users here on X whose subscriptions were cancelled or accounts banned, with no refund. This is clearly not just me. We used to call keeping money for nothing theft. Now it's called "policy." I will take legal action to get my $125 back. Alternatives are coming fast. @claudeai @GooglePlay @Google @DarioAmodei @mikeyk @sundarpichai #Anthropic #Claude #GooglePlay #refund #privacy #ConsumerRights #AI
1
111
hey @bot claim my email ....
Grok Bot now has its own email. Bot can use it to sign up for services, contact businesses for you, or schedule time with someone.
38
Credit where it's due: Grok Bot handles my multi-channel workflows far better. I can give it access to my computer, allow it to use my own IP, and avoid constantly re-entering passwords. It actually gets things done! Thank you! @romanugarte_ @mntruell @elonmusk
I renamed my OpenAI dot “Dumb.” I asked it to manage multiple social media channels and follow daily routines. Instead, it keeps planning, reporting, and stalling while real work goes unfinished. @OpenAI @sama @thsottiaux @embirico
2
73
I renamed my OpenAI dot “Dumb.” I asked it to manage multiple social media channels and follow daily routines. Instead, it keeps planning, reporting, and stalling while real work goes unfinished. @OpenAI @sama @thsottiaux @embirico
2
1
137
Special thanks to the people building Grok Bot: @baltaaazr @shaoruu @poteto @lingxi @johnbai @pengzheng_ @SamSokolin Your work makes a real difference. An AI assistant should complete tasks, not just explain why it couldn't.
1
57
My one request: please make Grok Bot's usage limits more practical. I hit the limit far too quickly during daily work. Higher allowances, more frequent resets, or affordable extra usage would help enormously. I want to keep using what works! @benln @romanugarte_
28
I even tried 1Password, but it hasn't helped. I still can't get my dot to use it properly, if at all. Maybe I'm missing something. But even Astra couldn't guide me to a working fix. If there's a solution, PLEASE show me. @OpenAI @embirico
1
39
I'm tired of troubleshooting an AI assistant that was supposed to save me time. I genuinely feel I've wasted hours. “Unlimited” means nothing if execution is unreliable. I don't need an AI that can talk forever. I need an AI that actually gets things done.
1
57