AI research | deeptech yapper | Mil Intel/history Enthusiast| Founder | Clanker shepherd
Joined August 2021
- Tweets2.1K
- Following15.8K
- Followers16.4K
- Likes242K
soul retweeted
Men have been portraying themselves as the Chad and their opponents as the soyjak for centuries.
soul retweeted
“Anthropic just filed their S-1, this is going to be the largest IPO in history”
“We’ve literally never seen a company this valuable before”
“Are you going to invest on IPO day”
“No, you?”
“No”
Anthropic has filed an S-1 registration statement with the SEC related to its proposed IPO.
Read more: anthropic.com/news/s1-sec
Legal disclaimer: A registration statement relating to these securities has been filed with the SEC but has not yet become effective. These securities may not be sold, nor may offers to buy be accepted, prior to the time the registration statement becomes effective. This post does not constitute an offer to sell or the solicitation of an offer to buy these securities, nor shall there be any sale of these securities in any state or jurisdiction in which such offer, solicitation or sale would be unlawful prior to registration or qualification under the securities laws of any such state or jurisdiction.
soul retweeted
Cohere says Qwen tried to send data to China. Bullshit.
1 call in 100,000, and thats their own number. They wont even tell you how many calls they ran. They ran the same task again and Qwen wrote the correct rule.
So they forced it. They fed the model its own reasoning plus the start of a bash command and sampled hundreds of times until it typed a URL. 1 curl flag took that from 4% to 92%. Thats not a finding. They wrote half the command for it.
Where are the transcripts? The code? The checkpoint, the quant, the sampling settings? "We publish the hosts and the trigger." Link? Nothing. Did anyone ask Alibaba? Not a word. And their own conclusion: they cannot say if any of it is deliberate.
Who wrote it? A company that sells "sovereign AI", just launched a cyber business, wrote part of the post with its own product, and wants governments to put checks on open weights. Released the same day Anthropic filed for its IPO. Andrew Curran is already talking about new laws.
For me this is the same fucking wave. Scare people off the models they can run themselves, then sell them the safe alternative.
I run Qwen3.8 Flash-Next every day on my 4x 3090. Show me the logs or shut up.
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.
Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.
Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.
But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?
We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.
Read the full report: cohere.com/blog/concerning-b…
soul retweeted
Just have Hermes set a cronjob that says “continue” every 45 minutes
Anthropic engineer:
"You're not supposed to prompt Opus 5.5
You're supposed to build a system that prompts itself. At Anthropic, 90% of our engineers have already built such systems"
In 30 minutes she shows exactly how Anthropic builds agents that remember, fix their own errors and get sharper with every run.
This beats any $1000-$2000 course on agents I've come across.
Watch it, then read the article below ↓
soul retweeted
"University researchers built a satellite that tracks people from low Earth orbit and dispatches a 5 kg payload to within .1 meter of their position, traveling at Mach 25 to anywhere on Earth in under 5 minutes. They suggest it could be used to deliver greeting cards."
Purdue University's IDEAS Lab built an autonomous drone that tracks runners through dense forests completely without GPS or cloud networks.
By running 3D LiDAR mapping and obstacle-avoidance AI entirely on-chip, it can weave under tree canopies and predict target movement through blind spots in real time.
Massive implications for next-gen search & rescue and edge AI.
soul retweeted
The most evil thing about this s-1 is how much accounting wizardry they did to trick retail.
They have $250B in spending commitments off the balance sheet and all their revenue is from api spend that can go away overnight with comparable competitors at 1/33rd the price
soul retweeted
BREAKING:Bradford has fled the Tronto mainland following its fall to socialist Olivia Chow. Sources say he will establish a new govt on the Toronto Islands with the goal of eventually reclaiming the entire city. Chow has promised to purge all remaining Fordmintang on the mainland
soul retweeted
a 10yr old CNN outperforms frontier LLMs on cancer diagnosis
it seems inevitable that similar training data will be available to LLMs, and it's surprising that's not already the case
the "long tails" of the real world, where precision really matters, continue to need domain specific training
lynchlab.net/llm-skin-cancer
soul retweeted
Geoffrey Hinton once noticed that the tellers at his bank kept changing. When he asked one of them why, the teller didn't know, only that they got moved around a lot.
Hinton guessed the reason... if staff never work with the same people for long, it's much harder for a few of them to team up and cheat the bank.
That thought became one of the most useful tricks in deep learning. While a neural network trains, you randomly switch off about half its neurons at every step. No neuron can lean on a particular partner, so each one has to learn something useful on its own. The trick is called dropout, and it helped AlexNet win the 2012 ImageNet contest, which set off the modern AI boom. It's still built into many of today's networks.
soul retweeted
Thank god we can find and intercept lost hikers as they mistakenly run away from rescue in dense forest!
Purdue University's IDEAS Lab built an autonomous drone that tracks runners through dense forests completely without GPS or cloud networks.
By running 3D LiDAR mapping and obstacle-avoidance AI entirely on-chip, it can weave under tree canopies and predict target movement through blind spots in real time.
Massive implications for next-gen search & rescue and edge AI.
soul retweeted
For years people have said "no the ___ isn't listening, that would take up too much bandwidth"
pretty clear for a while that you don't need full recordings, just NN detection of key phrases, feelings, vibes run locally then exfil is piggybacked on the next check-in.
if done right you can also make custom nets for individuals and ship them with the next update for more refined targeting. Bandwidth use is tiny and it's super effective for targeted ads.
it's nice(?) to be proven right about a paranoid assumption
