Democratization of AI-->Just found this video that I made about 9 months ago and it's oddly VERY applicable TODAY, perhaps even more so. DeepSeek just lowered the boom on the new Western AI companies' attempts to overprice AI and implement "usage limits" that price the average person OUT of the benefits of AI's wonders. Now, the Western AI companies simply CAN'T keep their ridiculous pricing structures or people will simply turn to options like DeepSeek. China has, quietly, yet again saved AI (hard to believe I'm saying this).
#DeepSeek #AI #AIDemocratization
Nous Portal really IS great. After nearly 2 months, I can attest that the Nous Portal subscription along with Hermes Agent is one of the easiest and most flexible options for autonomous AI use. Their new "top off" options have really closed the gap between $20/month and $100/month plans. And there is a free tier with excellent models to help you test it out. Give it a try, you won't be sorry.
One on One: Dr. J vs. Larry Bird--One of the best games of the time (for the Commodore 64). Back when EA was a smaller company!
#EA#EASports#Basketball#RetroGaming
LOL. This is called corporate GASLIGHTING. There have been MASSIVE layoffs in the field and new entry jobs are zero-nil. Spin much Jensen?
#AI#AINews#JensenHuang#Gaslighting101
Nvidia CEO Jensen Huang just called the idea of AI replacing software engineers "complete nonsense"
"The number of software engineers is actually increasing... if you can hire a software engineer and generate $9 trillion worth of productive work, why wouldn't you want to hire more?
If that line were flat, then obviously people would hire fewer software engineers. But because the output is so incredible, people want to hire more
This is going to show up in our economy somehow, soon. Useful AI has arrived"
Nvdia Keynote Summary:
NVIDIAโs highly anticipated keynote electrified the tech community, unveiling a massive, full-stack pivot centered squarely on the era of "Agentic AI". Leading the hardware charge, NVIDIA detailed its next-generation Vera Rubin computing platformโfeaturing a brand-new Vera CPU and a 10x leap in inference efficiency over Blackwellโspecifically engineered to handle the massive reasoning and memory demands of autonomous systems. For deskside local development, the company highlighted the DGX Spark and DGX Station hardware lines running alongside NVIDIA OpenShell, an advanced secure runtime built with strict guardrails to sandbox highly capable local agent workflows. On the consumer laptop front, systems powered by the premium RTX Spark hardware framework are slated to drop later this year, giving developers a highly efficient, unified architecture optimized for local AI pipelines. To fuel these complex agent environments, NVIDIA also dropped a major software bombshell with the release of the Nemotron 3 Ultra modelโa staggering 550-billion-parameter Mixture-of-Experts (MoE) powerhouse designed from the ground up to deliver frontier-tier logic to long-running, always-on enterprise agents.
#AINews #AI #Nvidia #JensenHuang #NvidiaKeynote
it's about even for me. i'm not a "view" chaser...so i comment only naturally...and my posts are PURELY for things i'm interested in that i think MAYBE someone else might be. if not? so be it. what's the point of chasing views/followers vs. just ENGAGING and having fun, interesting discussions?!
I want to say this about @NousResearch and @Teknium -- they RESPOND, they IMPROVE and they make a product that has literally taken the world by storm. I'm not hear to "glaze", I'm just giving credit where credit is due. Some of the things my agents are doing are hard to even believe (even if I showed the receipts) and it's all because of the Hermes environment.
Here's an interesting "tip"--Ask one of your agents "What skill do you think would be most helpful or interesting for you to create?" and LET them. Just try it...and watch! <---Can't do this with a frontier model via any front-end means. The novel customization opportunities are insane!
@Teknium@arvin17x Teknium, you do NOT have to defend your choices. You've created an amazing product my friend. With something like this, it's always gonna be a "damned if ya do, damned if ya don't" type thing. But look at the overall feedback? World-changing! THAT is what matters!
brad, the way i have my hermes set up? yeah, there is a system administrator component that is anxiety-provoking to say the least. pretty much every other update i have to "sleuth" issues and fix them. BUT it's worth it. i honestly don't know what i'd do without my hermes at this point and i was a BIG skeptic prior. stick with it. but lemme tellya, you DO kinda have to have a frontier AI model be your sidekick, even WITH the tips i gave you the other day. i hearya man!
It's been a very difficult ride. For over 2 months Grok went from my daily driver to something I use to verify X-posts. I get xAI is in a compute crunch, I get that SpaceXai's IPO is looming and they don't want to ruffle any feathers. But MAN this has been a tough go. While i still hold out belief that once the 1.5T param model is released, things will get better...but I have to admit the fear/worry they WON'T get better exists.
AI SECURITY ALERT! Rocco hacks Alexa and causes DEVASTATION!
LOL! I found this old video I made last year and had to share it. Hope you enjoy the levity!
#AISecurity#RoccoTheParrot#Alexa#AmazonAlexa
Joe Montana rookie card! I had EIGHT of these when I went off to college, sold 3 of them to fund my first semester spending needs. Later on, the 5 I had left would have been worth enough (for a period) to buy a small house...but of course, my mother accidentally threw them away. One of the most valuable cards of its era.
#FootballCards #Topps #TradingCards #JoeMontana
brad, i ALWAYS specify. one thing i've done is i give each agent their OWN directory on my C: and D: drives. under that directory, i create sub-folders for each output job. they religiously follow the filepaths. again, at first? there my be a hiccup for a job here or there first time or two...then? smooth sailing.
AI agents can now start real businesses.
@Meow lets your AI agent form a company, get an EIN, open a bank account, issue cards, and move money in a single prompt.
Works with Claude, Cursor, ChatGPT, Gemini, and more.
The first end-to-end company builder for agents.
don't forget Benjamin, SpaceXai IPO is looming. they do NOT want to make ANY waves...so we have to deal with a semi-lobotomized grok. it's frustrating as hell. but the HOPE is that, once the IPO goes through and the 1.5T model is released, they'll free up some compute and lower these ridiculous guardrails and re-routing to fast, cheap, light-on-compute models! fingers crossed my friend!
The official handoff has quietly begun.
I stated in my last "Mythos Threat" video that for human civilization to survive, the keys to our infrastructure will HAVE to be turned over to autonomous AI.
Humans are simply too slow and error-prone to fight machine-speed exploits. Google's announcement of an autonomous scanning, reasoning, and self-patching AI immune system proves it.
Bad AI attacks in milliseconds.
Defensive AI repairs in milliseconds.
Humans are completely removed from the loop.
This is monumental, yet most will completely miss it. The keys are being turned over. This playbook will scale to every sector of life very soon.
https://t.co/Nh5HTHKJPE
#AISecurity #Cybersecurity #GoogleGemini #AutonomousAgents
Today weโre introducing Google AI Threat Defense - a comprehensive AI-powered cybersecurity solution designed to help continuously monitor for and stop AI-powered threats before they can impact your business. Hereโs how it works:
1. AI Threat Defense uses our cybersecurity platform @wiz_io to scan and prioritize what applications and systems have the highest security risk.
2. Gemini and other frontier AI models can then autonomously perform continual deep scanning of your applications - starting with those at the highest risk - to identify security vulnerabilities.
3. The capabilities of CodeMender - a new software repair agent - are then used to verify and accelerate the patching of vulnerabilities.
4. And our @wiz_io autonomous agents continuously test your systems to find unknown vulnerabilities before adversaries do so that you can remediate them before you are attacked.
While other model providers focus on using AI to find and flag vulnerabilities, Google AI Threat Defense actively prioritizes your most critical real-world risks and accelerates their remediation using a variety of models since no single model finds a superset of the vulnerabilities found by all other models.