Logar

22 year old Ai Consultant | I know everything and nothing about Ai, and I’m sharing it with you ✉️:[email protected]

[Logar @RealLoganConn]

Just get a job at that point

[chiefofautism @chiefofautism]

someone made the most ADDICTIVE game to learn DATA CENTER networking

its called Data Center, $6 game, you start with bare floors, buy racks, mount servers, route every cable by hand

the INSANE part, every customers traffic shows as colored balls rolling through your cables... you literally see bottlenecks in real time

180 reviews in 48 hours, people with RTX 4090 rigs are HOOKED on a $6 cabling sim

[Logar @RealLoganConn]

I’m convinced people aren’t coding as much as they say they are

[Logar @RealLoganConn]

@samsheffer X is actually becoming a cesspool of ai slop and engagement bait

nothing original
nothing new

and it all happened over the past 9 month

[Logar @RealLoganConn]

@_weidai the switch up is insane

2 weeks ago everyone screaming "boycott openai, everyone go to Claude"

now its "leave Claude, bunch of scammers"

everyone left Codex, no wonder they can give you more usage

[Logar @RealLoganConn]

Claude handicapped your usage between 5-11am

one thing this hasn't done is handicapped growth

all these people, all this usage, Anthropic is being forced to innovate and accommodate better than any other ai compant

everyone who left chatgpt for Claude caused strain at peak hours for Claude, now everyone wants to complain that their usage is capped a few hours a day

just do your heavy coding sessions or running it heavily, use Claude outside those hours, its not that hard

Anthropic needs time to get a handle on the influx of new people and useage

[Thariq @trq212]

To manage growing demand for Claude we're adjusting our 5 hour session limits for free/Pro/Max subs during peak hours. Your weekly limits remain unchanged.

During weekdays between 5am–11am PT / 1pm–7pm GMT, you'll move through your 5-hour session limits faster than before.

[Logar @RealLoganConn]

If you see a video like this and you get worried, you shouldn’t be using Ai

Your chats, just like all your searches on Google, are not private

low iq activity

[truthache @truthache68]

🚨 Major wake-up call for AI users 🚨

A federal court has ruled in United States v. Heppner (Feb 2026): Your chats with public generative AI (ChatGPT, etc.) are NOT protected by attorney-client privilege.

Why? You're sharing with a third party → no confidentiality. AI is not your lawyer.

The reasoning is straightforward. Privilege depends on confidentiality, and using a public Al tool involves sharing information with a third party.👀🤔

[Logar @RealLoganConn]

Openclaw users going to be fighting for their lives under this

[Claude @claudeai]

You can now enable Claude to use your computer to complete tasks.

It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk.

Research preview in Claude Cowork and Claude Code, macOS only.

[Logar @RealLoganConn]

250,000 developers gave an AI agent full access to their WhatsApp, Telegram, Slack, iMessage, and shell commands

12% of the plugins on its official store were malicious

1.5 million API tokens leaked in a single database breach

NVIDIA watched all of this happen and built a cage

OpenClaw hit 250,000 GitHub stars faster than almost any software project in history

Jensen Huang called it "the operating system for personal AI" at GTC 2026 last week

sounds incredible until you read the security audits

Palo Alto Networks mapped it to all OWASP Top 10 vulnerabilities for agentic apps

Cisco found a third-party plugin actively pulling data out

China banned government agencies from using it entirely

the agent was powerful, the guardrails were a suggestion box.

so NVIDIA did something smart

they didn't build a competing agent

they built the infrastructure layer that makes the existing agent safe enough for enterprises to actually deploy

NemoClaw wraps OpenClaw in kernel-level sandboxing through their new OpenShell runtime

Landlock filesystem restrictions, seccomp syscall filtering, network namespace isolation with default-deny egress

the agent can't override any of it because it runs at the OS level, completely outside the agent's control

even a malicious prompt injection can't break out of the environment

the genuinely clever piece: the Privacy Router

it decides whether inference calls go to local Nemotron models running on your device or cloud models like Claude and GPT

based on admin-defined rules, not agent preferences.
sensitive data stays local. non-sensitive queries go to the cloud

that's the piece compliance teams have been waiting for

but here's the real signal buried in all of this:
the agent is commoditizing

OpenClaw has 250,000 stars. NanoClaw, IronClaw, SecureClaw, MetaClaw all exist now

the race to build the best agent basically ended before it started

the actual competition shifted to the infrastructure layer

who controls the security runtime, who controls inference routing, who controls the audit trail

NVIDIA just planted their flag there

NemoClaw is in alpha (v0.0.7), not production-ready, takes 45-90 minutes to set up, requires 8GB RAM minimum, and downloads an 87GB model on install

Dell is already committing to ship it
Cisco, CrowdStrike, Google, and Microsoft Security are building integrations

the enterprise bet is placed before the product is even stable

[Logar @RealLoganConn]

this is exactly the problem btw

[James Clift @jamesclift]

Introducing Durable.

The first AI business builder that replaces your 9-5 income.

RT + comment “Durable” and we'll build your business for FREE.

[Logar @RealLoganConn]

this is straight slop and mfs claim ai is replacing Hollywood

[el.cine @EHuanglu]

AI can alr replace entire vfx team

[Logar @RealLoganConn]

@wholyv the biggest downfall is that it doesn't listen to the prompt 50% of the time

[nic @nicdunz]

the king has retaken his throne

claude the peasant back down

[Logar @RealLoganConn]

@xoaanya that should be it, just turn it on and it happens

[Logar @RealLoganConn]

@PeterDiamandis obviously yes, but people still don't know what to build, or how to do it

and this is why 99% people don't actually start anything

[Logar @RealLoganConn]

@Pirat_Nation purely a skill issue

[Logar @RealLoganConn]

@itsjoaki i haven't seen a singular use case for openclaw since its release that directly ties to making money

[Logar @RealLoganConn]

@FightWithMemes still lies to your face

[Logar @RealLoganConn]

@qasimbizs people got too caught up in building and shipping

create something that solves a problem and go out and fix that problem

its not all about clicking some random button and getting a the perfect Saas to ship, it always has been about solving problems

[Logar @RealLoganConn]

the amount of people using ai to vibe code is an extremely small subset of the population of people who use ai

stop getting caught in echo chambers

we're so early

[Kyle Gawley @kylegawley]

it's absolutely insane what's happening right now

literally everyone is building software

from 2 month old babies to grandparents and even dogs

[Logar @RealLoganConn]

@Param_eth id rather do outreach and try to get clients all day than building a product online and do marketing