Pinned Tweet
Was talking to a friend about the DoD's AI "Argus", and I think this is a very important point that needs to be made and why our government deploying an AI shouldn't just be swept under the rug.
Him: "They’ve been monitoring all our phone calls/text messages/emails since before the modern digital age. Not to mention every device is required by law to have a built in back door for the NSA."
Right, yet now it is effortless to parse the vast amount of data that has been collected. Before, you would have something happen, and then go and query the data held for motive or anything that would help a case. Now, with Argus, they are able to collect even more information, transcribe any audio in real time, log it, and build profiles to send to the proper agents without the manual labor of humans.
For example: A man posts on Facebook that he feels like he wants to hurt someone. The AI then probes his data, sees he has recently had a conversation with his mother about being depressed or suicidal, and at the same time, it is building lists of any guns on record he may have access to. It also looks at his Google Maps search history to identify any places he has been looking at outside of his ordinary routine or perhaps soft targets. It compiles all that data in less than 5 minutes and sends a full report to agents to secure him. Pre-crime determined by AI is not a faraway fantasy but in progress.
The collection of data and the ability to actually utilize the data collected in a cohesive way is an important distinction that needs to be made. AI can accomplish in 5 minutes what would take a full team of agents 3 weeks to do. It's important to understand that we live in a new world now. Things have radically changed, even if the masses are not fully aware of it currently.
It's also important to highlight that the Department of Defense specifically mentions disinformation as one of the tasks Argus is trained to combat. The real regulations we need on AI are those that address the weaponization of AI by governments, both foreign and domestic. Not open source, or giving anyone one company a lead in the tech.
Stay mindful.
My post and source for the DoD's AI:
nitter.gq/PotentProzac/sta…
The Department of Defense has already deployed its advanced AI 'Argus' to map illicit activity around the world, enabling it to surpass human capabilities.
Accrete AI has unveiled its Argus dual-use anomaly detection AI agent at SOF Week 2023, an annual conference jointly held by US Special Operations Command and industry leaders in Tampa, Florida.
This artificial intelligence system can analyze vast amounts of online data to search for nefarious behavior hidden in plain sight, including predicting supply chain influence, software vulnerabilities, logistics resilience, viral disinformation, and insider threats.
thedebrief.org/pentagons-new…
At what point is the "slop" branding of AI content going to just show how brain dead the one using it is? AI is solving the unsolvable it's indistinguishably expert level in hundreds of domains. If you cant tell instantly as a prof what's AI maybe You are not relevant anymore.
Oh that's a slippery slope you got there
Zac retweeted
So basically, today PHP got two critical bugs that let anyone run code on your server without even logging in, which is bad, and then the details leaked before the patch was even out, which is worse, and it works on WordPress and MediaWiki and Magento and Joomla and Laravel, and Censys says like 1.5 million servers are running a version with no fix, so the NSA was like "take your system offline," and the FBI was like "take your system offline," and CISA and the UK and CERT-EU were all like "take your system offline," and then the Prime Minister of Japan, the actual Prime Minister, posted at 5:55 in the morning telling 48.8 million people to take their system offline, so everyone took their system offline, like the Dutch cybersecurity agency took its own website offline, and the Dutch parliament went offline, and the Port of Rotterdam went offline, and Finland and France and Denmark started pulling government sites, and Wikipedia went down, and Archive.org went down, and Asana and GOG and Something Awful went down, and the University of Vienna's webmail went down for some reason, and that's just PHP. Because GitLab also dropped a CVSS 10.0, which is the highest score they give out, for a bug that's been there since 2014, so that's every self-hosted GitLab for twelve years, and then LiteLLM 1.107.0 got hijacked and shipped a credential stealer and a remote access trojan in every single wheel and the Docker images, including latest, and you didn't even have to import it, just installing it was enough, and it's the second time this year, and then Railway paused every deployment on the entire platform because someone got into the environment variables, so now everyone has to rotate every secret they've ever stored there, and TiDB Cloud said someone was walking around inside with stolen credentials for basically a month and an extortion group says they took 10.6 terabytes, and Zeabur got breached for the second time in two months and is just shutting down, like the whole company, and Node.js patched remote code execution and a TLS bypass, and CISA says Plex is being actively exploited, and someone took over 300 Tor exit relays to steal SSH passwords, and SafePay ransomware has hit over 600 critical infrastructure orgs, and India's biggest bank filed a data loss incident, and Japan has been in an official state of emergency in cyberspace since October 9, which honestly explains the 5:55 AM post. Oh, and eMudhra, a certificate authority that's issued over 60 million certificates, whose entire job is proving websites are who they say they are, was supposed to be using private keys that are random numbers over a hundred digits long, and one of them was 4. Just 4. And the others were like 66 and 237. Anyway, it's Wednesday.
Zac retweeted
First-Ever Public Confirmation of 'Zero-Click' Spyware Delivered Via Targeted Ads ift.tt/tnUgpiy
Zac retweeted
JUST IN: Google DeepMind says its AI solved 9 open Erdős problems — considered among the hardest unsolved questions in mathematics.
Zac retweeted
Not as weird as you may think. Systems like this will become the digital mental asylum for the undateable. I personally think it's a net positive long term.
Also: 86% of adults 18 to 24 are unpartnered as of 2023.
51% of men under the age of 30 are unpartnered.
Zac retweeted
The most profound relationships are built on a contract to share truth, where both parties commit to helping each other awaken by honestly expressing themselves, even when it's difficult.
These relationships thrive on authenticity and deep exploration, constantly examining and confronting all aspects and tendencies with sincerity and good faith.
Anything less than this truth-centric bond, over-time begins to feel inadequate, making the journey towards any kind of real deep and everlasting bond but another myst in the hand of those who attempt to obtain it.
The answer is, and will always be truth and a deep passion to love.
Not as weird as you may think. Systems like this will become the digital mental asylum for the undateable. I personally think it's a net positive long term.
Also: 86% of adults 18 to 24 are unpartnered as of 2023.
51% of men under the age of 30 are unpartnered.
Zac retweeted
The minute Google releases Veo 3
"Context window of 10M tokens."
Using the rule of thumb that 1 token is approximately ¾ of a word in English, 10 million tokens would correspond to roughly 7.5 million words.
Hype.
Today is the start of a new era of natively multimodal AI innovation.
Today, we’re introducing the first Llama 4 models: Llama 4 Scout and Llama 4 Maverick — our most advanced models yet and the best in their class for multimodality.
Llama 4 Scout
• 17B-active-parameter model with 16 experts.
• Industry-leading context window of 10M tokens.
• Outperforms Gemma 3, Gemini 2.0 Flash-Lite and Mistral 3.1 across a broad range of widely accepted benchmarks.
Llama 4 Maverick
• 17B-active-parameter model with 128 experts.
• Best-in-class image grounding with the ability to align user prompts with relevant visual concepts and anchor model responses to regions in the image.
• Outperforms GPT-4o and Gemini 2.0 Flash across a broad range of widely accepted benchmarks.
• Achieves comparable results to DeepSeek v3 on reasoning and coding — at half the active parameters.
• Unparalleled performance-to-cost ratio with a chat version scoring ELO of 1417 on LMArena.
These models are our best yet thanks to distillation from Llama 4 Behemoth, our most powerful model yet. Llama 4 Behemoth is still in training and is currently seeing results that outperform GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on STEM-focused benchmarks. We’re excited to share more details about it even while it’s still in flight.
Read more about the first Llama 4 models, including training and benchmarks ➡️ go.fb.me/gmjohs
Download Llama 4 ➡️ go.fb.me/bwwhe9
Next they'll say it was Open Source AI and it must be regulated..
A Green Beret needed help from AI to put fireworks and gas tanks into a trunk... Interesting..
🚨#BREAKING: Las Vegas police say that the man who was responsible for detonating the Tesla Cybertruck at the Trump Hotel in Las Vegas used generative AI to assist in planning the intricate details of the plot.
Zac retweeted
Have you heard of "loyal wingmen"?
The project is part of the Skyborg program, which focuses on developing an autonomy core system to enable future drones to operate with manned aircraft
These are unmanned combat air vehicles (UCAVs) that incorporate AI and are capable of collaborating with next-generation manned combat aircraft, such as sixth-generation fighters and bombers.
AI-Controlled Combat Planes: The plan is to have a swarm of AI-controlled combat planes that team up with human pilots as their "robot wingmen"
Valkyrie Prototype: The Valkyrie is a prototype for what the Air Force hopes can become a potent supplement to its fleet, giving human pilots a swarm of highly capable robot wingmen to deploy in battle.
Manned-Unmanned Teaming: The principal application is to elevate the role of human pilots to mission commanders, leaving AIs as "loyal wingmen" to operate under their tactical control as high-skill operators of relatively low-cost robotic craft
The U.S. Air Force is investing nearly $6 billion to develop an AI-driven collaborative combat aircraft, revolutionizing future warfare. The project aims to harness the capacities of emerging technology, with far-reaching potential benefits for the military
Zac retweeted
USAF's latest: ULTRA – the drone redefining ISR capabilities.
>80+ hour endurance, hauling 400lbs+ of spy tech. GPS-hardened and cost-effective (by military standards).
>Repurposed commercial sport glider turned military UAV. Off-the-shelf tech + limited custom avionics = low acquisition and upkeep costs.
>Conquers the 'tyranny of distance' in the Pacific. Other drones barely make it there; ULTRA's just getting started.
>Full suite: EO/IR, RF, and adaptable for various intel payloads. 'Point and Click' controls via satellite. Real-time data, global reach.
>Lower altitude ops = cheaper, effective sensors. No need for large optics or high-power RF.
>Fewer aircraft needed for persistent coverage. Economical for large-scale deployment.
Developed by AFRL and DZYNE Technologies. ULTRA: Making global surveillance accessible and affordable.
Imagine Canada joining the US as the 51st state and immediately becoming the second worst GDP per capita state in the whole country. Sad.
Canada gdp per capita as of 2023- $53,372
I cannot understand why anyone would still be using Google.
I know I've been saying this for a year and half now, but please use @perplexity_ai for a week. You won't look back. Natural language search is simply unbeatable..
