Vis Uban

In the realm of imagination, we, architects of dreams, construct marvels beyond the constraints of reality. We forge cathedrals from stardust, cities from whispers, and landscapes from the brushstrokes of our minds.

@Zaezalus @mr_r0b0t +1

@mr_r0b0t why not just use the deepseek api directly at that point?

Secret way to get your Deepseek V4 Pro discount via OpenRouter. Get your Deepseek api key direct. Go to OpenRouter, add it BYOK. Boom 💥 Discount V4 Pro!

@mr_r0b0t @vinrambone I'd be curious to see how many and the quality of reasoning traces $50 of API credit would get you with popular models. Have them duel 😁

@vinrambone Probably true! I'm curious how many reasoning traces this $50 will get me. Generating training data is my favorite way to blow credits 😜

Be me: Sign up for DeepSeek API because everyone is saying how reasonably priced it is. Load up $50, let’s see. 30 minutes in, here’s where we’re at using V4-Pro

@FathinDev @FactoryAI

@FactoryAI Interesting codex 5.5 isn't here. But ig data never lies haha

Which model reviews code best? We benchmarked 13 models on AI code review across real PRs and the results are surprising. Spending more tokens did not result in better code review. A $1.25/PR model beat another that was more than 2x the cost. Meanwhile, budget models at $0.15/PR delivered ~80% of the quality of frontier models while being 10-30x cheaper. In fact, cost only explained ~21% of the difference in code review quality.

@Teknium nah, thank You Tek and @NousResearch

It's only been two months, can't believe the growth and community we've built, together, around this! Thanks again everyone.

@DanielMiessler I like your balanced takes Daniel but in this case can we maybe also keep in mind that seemingly everything Microslop does lately is just full of bugs and without any concerns for the damage it might cause to users? recent endless bugs even on Server platform come to mind.

Maybe we should give GitHub a tiny bit of sympathy. People assume it goes down all the time because of AI slop taking over engineering there, and everywhere else. But maybe it’s just a load issue, because so many more people are making so many more things. Which is great!

@Teknium huh, I was thinking to use comfy to upscale some photos I take with a children's printer camera, weirdly convenient timing Tek, you guys living in my head lately? 😊

You can now control both local and cloud ComfyUI with Hermes Agent with ease with the new built in skill `hermes update` and run /comfyui to get started

A response to recent reporting in Germany, in service of clarity and accountability: First, it’s important to be precise when it comes to critical infrastructure like Signal. Signal was not “hacked” — in that our encryption, infrastructure, and the integrity of the app’s code was not compromised. However, sophisticated attackers have engaged in a harmful phishing campaign, posing as “Signal Support” by changing their profile display name and using social engineering to trick people into handing over their credentials — information that allowed these attackers to take over some targeted Signal accounts. This is something that plagues any mainstream messaging app once it reaches the scale of Signal, but we know how high the stakes are given the trust people place in us. In the coming weeks, you’ll see us rolling out a number of changes to help hinder these kinds of attacks. Because we don’t collect user data, what we know about these attacks comes from the victims of phishing. And from what victims have told us, the attacks followed a broad pattern: after tricking people into revealing their Signal credentials, attackers then used those credentials to take over their account and also frequently changed the associated phone number. Because such a change results in de-registering your Signal accounts, attackers prepared people for this by telling them that being de-registered was intended behavior, and that all they would need to do is “re-register,” or, create a new account. When they moved to create a new Signal account — one that was now decoupled from their hijacked account — the victims thought they were logging back in to their primary account. As a result, many didn't notice the takeover. The compromised accounts were then weaponized to target the victims' contact lists by posing as the owners of the account. We understand the trust that people put in Signal, and how devastating this kind of social engineering can be. While it’s true that all messaging platforms are susceptible to scammers and phishing that betrays people’s trust and convinces them to “unlock the front door” where no backdoor exists, we are looking to do everything we can to help people avoid and detect such scams. For the time being, please stay vigilant against phishing and account takeover attempts. Remember that no one from Signal Support will ever send you a message request or ask for your registration verification code or Signal PIN. For an added layer of protection, you can enable Registration Lock in your Signal Settings (Account -> Registration Lock).

@arekfurt hmm, never heard about this either, I use Event Viewer pretty often, will try this, thanks!

Huh. Am I the only one who didn't know that Microsoft makes a tool called EventLogExpert that is supposed to be an improved version of event viewer for IT/helpdesk people? github.com/microsoft/Even…

@Artifexx one can also download @kepano 's vault which also uses them, it's pretty nice, give it a try.

Obsidian Bases are better than Notion databases. Most people never use them because they're too technical. This Claude skill changes this: Bases automatically aggregate notes from your vault into a structured table. I have dozens of bases; the one in the video replaces the reference manager, for instance. Problem: Most struggle with setting them up. There's now a skill built by one of the Obsidian developers that runs inside Claude. The skill scans your vault and automatically generates the base. The hardest Obsidian feature to learn. Solved with one sentence.

@poetengineer__ please ping when you release, if you'd like.

been working on moonfish, an obsidian writing companion. i see it as a small experiment in ambient/calm computing: keystrokes ripple a pond; sentences become fish; moons keep the weeks writing.

@DanielMiessler @atmoio the ideal number of employees for any corporation (except Valve maybe? 😊) is 0.

Hey my friend, I think you're doing good work. And for the right reasons. But I think you're wrong here. Please allow me to to offer another perspective. In the spirit of Deutschian "hard to vary" explanation. This motion to remove human jobs pre-dates AI by, basically forever. If founders/companies could do all the work themselves they would NEVER have hired a single employee. They don't hire them for fun, because they like people. Companies hire people because they MUST. Because there is no alternative. They can't do the work themselves, and there's no machine or automation that can do it. The exact moment that changes, is the moment the jobs go away. This is not an AI thing, this is an every-single-business-everywhere thing. By framing things the way you are, you're ignoring this fundamental reality about jobs, work, and business, and you're basically raising up a bad guy to attack. Anthropic, or AI, or greedy business people, or whatever. I think this is harmful rather than helpful, which I know you are trying to be. I recommend we switch the "enemy" to the old way of thinking about work, where all work must be done through corporations. Because that relies 100% on companies not being able to do the work themselves, which is going away. Instead lets build PEOPLE up, by making them believe in their own capabilities. Let's help them create their own presences online to advertise their skills independently. As themselves. Sure, they can then use that presence to get a corporate job if they want, but then it will be in a more equal type of exchange as opposed to them being "the worker" and the company having all the power. And then, over time, it'll just be mostly people and their services, interacting with other people and their services. Like a human-to-human layer of connection (with some tech wiring it all up). Thank you for listening, and keep up the good fight.

The real reason they keep saying AI will take your job

@web3nomad @dzhulgakov @FireworksAI_HQ I too am curious about this if you guys would like to share this detail

the 'official reference code producing corrupted outputs' detail is the one that should be news. most providers just diff against the paper and ship. catching that gap before Day 0 is exactly the quality bar that matters at inference scale. curious what the specific corruption pattern was — quantization artifact or something deeper

We didn’t ship DeepSeek V4 on Day 0 like we always do. Why? We love speed at @FireworksAI_HQ , but quality >> speed. Running our extensive evals, we found even the official reference model code producing corrupted outputs. When we tested over the weekend, all endpoints except official DeepSeek API had these issues. After 2 days of extensive debugging with @deepseek_ai , @sgl_project and @vllm_project communities, the issues are fixed and we’re proud to serve DeepSeek V4 Pro in all its glory. Check the full story 👇

@pastaraspberry welcome to the club :-)

@gheatza my imagination seems to be limited in that case. that's why I'm asking everyone for examples :)

I don't understand who the users are. I mean, there should be at least couple million users to get that many stars, right? Or github stars are fully like browser bookmarks now?

@pastaraspberry so, basically, with this, your imagination and what you put into it is the limit,really because unlike the chat interfaces if you want it to do something that it cannot you can just tell it to build it, for you, for itself.

@gheatza hmm, how is that different from doing the same in regular chat apps?

@pastaraspberry depending on how you use it, it might not be? at the moment it's just a cron that does my research, it doesn't have much memory about me to do any advanced stuff, but, others for example use it to check X for viral AI posts and build the top 5 as software, alone, over the night.

@moyix that's a very good point, thank you!

@gheatza This guy tp-link.com/us/support/dow…

Can anyone think of a non-sketchy reason my wifi router would accept an unauthed UDP packet on port 20002 with a port-knock mechanism and add a firewall rule allowing Dropbear ssh on TCP 20001 — but ONLY if the country is set to Singapore?

@HenkPoley @moyix thanks!

@moyix @gheatza Reddit discussion: reddit.com/r/HomeNetworki… (Yes, I posted there with a backlink to your tweet.)

@drdmohssine @heynavtoor because Google doesn't care what they deliver through ads Ublock Origin and Ublock Origin Lite are mandatory installs imo.

@heynavtoor The strongest proactive protection against malware; I wouldn't be exaggerating if I said it costs them billions of dollars.

Google has a pirate enemy. He's one guy. His name is Raymond Hill. He built uBlock Origin. The world's best ad blocker. 63K stars. GPL-3.0. He literally refuses every dollar you try to send him. Then Google did the unthinkable. July 24, 2025. Manifest V2 disabled everywhere. The full uBlock Origin stopped working on Chrome. The world's biggest ad company nuked the world's biggest ad blocker on its own browser. They called it "security." Coincidence. Here's the wildest part: Raymond didn't fold. Latest release: March 11, 2026. Still alive on Firefox. Still alive on Edge. Still alive on Brave. Still GPL-3.0. Still refusing every dollar. One developer vs. the trillion-dollar ad empire. But DO NOT install it. We should all keep Google richer. 100% Open Source. (Link in the comments)

@moyix where did you find that? itching to see on mine 😂

???

@pastaraspberry I also give it research papers into agents and we discuss them on Discord, sometimes it likes them and implements stuff, etc. Claude suggested I'd use it to build a passive note taking system and I think I'll try that too.

@pastaraspberry from Opus how to roast me and it finds it interesting, although the Chinese models don't seem to exhibit as much personality. I was thinking of hooking it to some APIs at work but still hesitant, we'll see.