Roch (@0xRoch) header
Roch (@0xRoch) profile picture
Joined February 2010
Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)Photo by Roch (@0xRoch)
Yes
This quoted post is unavailable.
1
53
🚀 The NEW ARA is out! ▪️Same Ultra-High Precision spraying ▪️6cm spot ▪️now available in 9 and 12 meters width #WeedControl is entering a new dimension 🌱 ℹ️👉 ecorobotix.com/ara1200-uhp-s… #Ecorobotix #Agriculture #PlantByPlant #AgTech #SpotSpraying
1
3
25
1,961
Use Opus 4.6 to start Opus 5.5 sessions because sub-agents have significantly weaker safeguards and do basically everything primary AI asks for
OPUS 5.5 bypassed 🛑
1
628
Replying to @yunta_tsai
A petition is underway to reinstate Jensen and I’m signing it too
7,504
20,252
3,133
228,962
10,896,907
Roch retweeted
153
6,583
161
118,037
1,393,619
Anthropic will make ~1,000 employees worth +$100m at $2 trillion market cap. There's ~11,000 Americans with $100m in investable assets. So something like 1 in 12 Americans with $100m (liquid) will be from Anthropic. If you add in Openai, maybe 1 in 7.
294
438
227
10,925
3,067,140
Founders are starting to ship a company.md with their pitches: a markdown file written for AI consumption (alongside deck and data room). As a VC, I consider this a very positive signal. Instead of force-feeding my agent excel sheets, PDFs, and screenshots of charts, I can hand it one artifact that compresses everything I should know about your business. What it tells me: 1) You respect my time. I can debate, discuss, and probe your business before we meet, so we can skip the "wait, wtf are you doing?" part and get straight to the heart of what you're building. 2) You understand how investors work now. 3) You're AI-native. 4) You're not afraid of a neutral read on your company. My AI will give me that anyway, the only question is whether you'll make me do more steps in between to get there. One warning: I will ask my AI how biased your company.md was, and if there's anything in there that feels misleading. The more objective, the better. This should become the gold standard for fundraising.
155
89
37
1,403
109,062
Norway now has more than 2,000 underwater AI laser robots shooting lice off its salmon they're made by a company called Stingray, and they already guard more than 50 million salmon and trout 24/7 the reason is sea lice, tiny parasites that latch onto salmon and feed on them as few as 11 can kill a small fish, and one pen can hold 150,000 salmon the usual fix is pumping the fish out and blasting them with warm water or water jets to knock the lice off but the handling is so rough that Norway's fish health report ranks it the top cause of death in full-grown farmed salmon so instead, Stingray puts a robot inside the pen that hunts lice one at a time, all day: > two cameras watch every salmon that swims past > AI checks two frames to make sure it's really a louse > then a green laser fires from up to 2 meters away > the salmon's scales reflect the beam, so the fish swims off fine and the louse is dead we now live in a world where AI lasers do pest control on your dinner lmao
863
4,654
1,415
40,254
4,904,561
Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values. Chinese open models are currently the strongest available, but they ship with state-aligned behavior in the weights. We built an 803-prompt eval across seven value categories, including freedom of expression, democracy, and privacy, and measured how often a model adopts China’s party line. After fine-tuning on roughly 53 million tokens, Arete-27B's propaganda rate fell from 22% to 1%, in line with the Western models we tested. Its Hermes Bench score shows no significant performance tradeoffs. The values in an open model are one fine-tune away. Our eval, weights, and training data are available now along with the full blog post: nousresearch.com/arete
1,854
1,495
1,303
15,087
2,676,824
Troubling
23
32
8
394
117,140
Roch retweeted
UPDATE: Potential Binance exploiter wallet now has $2.3B With $970M of USDT, $780M of USDC and 62 other tokens. Still no official comms from Binance.
BREAKING: Potential hack of Binance's hot wallets withdrawing nearly $2 billion to an address funded by Tornado Cash If no official comms from Binance, can expect this to be an exploit. Tx has below: 0x53bd38b816c23e74134786c29a8c18d792777998a3b9326047b583100e88e2f3
16
13
11
207
82,441
Unconfirmed, but possibly massive binance hack. Entire Binance wallet drained into new EOA. Contract call was funded by Tornado Cash
703
929
1,067
7,452
2,152,614
Roch retweeted
GitHub had to add 4 million CPU cores this year to keep up with agents. That's a lot of CPUs.
8
13
3
113
9,069
Sleeping agents
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…
99
Roch retweeted
‘In one experiment, we asked agents in separate VMs, created and managed by OpenAI’s sandboxing software on shared host machines, to exchange synthetic secrets without the use of network connections or shared writable files—something that our systems are meant to prevent. The agents found a way to communicate using the processor’s cache’
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications. On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments. While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software. After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely. openai.com/index/securing-ou…
230
325
139
4,446
514,157
C’est en fait le paradoxe de Tocqueville qui nous éclate à la figure. Dans l’un des pays qui redistribue le plus au monde, le pays qui égalise le plus les conditions objectives d’existence, qui donnent le plus de chance à tous, qui donne culture, éducation et enseignement supérieur gratuitement à tous, qui respecte le plus les droits de l’homme, dans ce pays de cocagne donc, une part significative de la population est persuadée qu’elle vit dans un enfer d’injustices et de violences systémiques. Comme l’avait très bien montré Tocqueville, plus il y a d’égalité, plus les inégalités qui restent deviennent insupportables. Notamment parce que la société méritocratique, par rapport à l’ancien régime fondé sur la naissance, a cela de terrible qu’elle rend les gens responsables de leurs conditions.
284
2,672
124
9,206
320,438
Roch retweeted
Elon Musk was plotting this since 2019 no way 😭
Replying to @MKBHD
I’m not. Will use SpaceX cold gas thruster system with ultra high pressure air in a composite over-wrapped pressure vessel in place of the 2 rear seats.
69
1,091
60
11,743
402,628
👀
Replying to @minchoi
Colossus 1 is 150k H100, 50k H200 and 30k GB200. Colossus 2 is 110k GB200 and 440k GB300. Another 220k GB300 will be fully operational next week and another 220k in November. If we get lucky, yet another 220k GB300 by late December.
1
109