cyr4x (@_cyr4x) profile picture

Cybersecurity/AI professional by day. Creative by night. https://teapaw.org/t.co/L0Nx7BIViR https://teapaw.org/t.co/sFXlumL42s

earth
Joined June 2014
Photo by cyr4x (@_cyr4x)Photo by cyr4x (@_cyr4x)Photo by cyr4x (@_cyr4x)Photo by cyr4x (@_cyr4x)Photo by cyr4x (@_cyr4x)Photo by cyr4x (@_cyr4x)Photo by cyr4x (@_cyr4x)Photo by cyr4x (@_cyr4x)
Pinned Tweet
Introducing SoundShredder - Local audio cleanup tool for creatives working with video/audio. - Presets to isolate voice, music, sfx - Prompt for sound removal - Isolates sounds into stems - Mixer control for your stems - More to come. Made for the #Higgsfield community. Available for Mac/Windows. Download here: github.com/xD4O/SoundShredde…
1
1
3
247
cyr4x retweeted
A lot of people are mocking "Western values" without reading the blog post. What Nous actually did is simple: They took a Qwen3.8-27B that gives CCP talking points on sensitive topics, fine-tuned it on ~50k human-checked examples, and that behavior went away. CCP propaganda answers went 22% to 1%. Please read before posting.
Oh wow!!! 😮 Nous trained their own Qwen3.8-27B and released Arete-27B 🔥🔥 This model stops following China's "guidelines" on things like speech, democracy, and privacy. This is big deal for anyone running Chinese open models locally! Great job @NousResearch I love it. I hope it's first of many!
234
41
38
889
148,378
. higgsfield.ai/community/room… Hey guys this my AI CINEMATIC VISUAL room . Join up , fresh content every day with @higgsfield
3
5
27
1,013
cyr4x retweeted
If you own 2 DGX Sparks, skip M5U 256 GB If you own 4 DGX Sparks, skip M5U 512 GB Have money to spare? - If you have 1 DGX Spark, add another - If you have 3 DGX Sparks, add another Want faster inference? Buy GPUs.
96
21
12
514
76,972
Jim Cramer is a legend
594
3,876
764
26,593
1,999,888
NVIDIA Nemotron 3.5 Transcribe rocks! I had it in early access and it's ultra versatile, give it a try!
Meet NVIDIA Nemotron 3.5 Transcribe 🎙️ An open-weight, promptable 30B MoE model with ~4B active parameters that brings transcription in 40 languages, speaker attribution, translation, contextual prompting and audio understanding into one checkpoint. Watch it keep track of who said what when everyone’s talking at once. 1/ 🧵
8
7
126
13,543
‼️ Anthropic and Google researchers found a flaw that lets one crafted packet hijack a server during network boot, before Secure Boot checks the OS. Almost all servers on earth are affected. CVE-2026-97631 sits in TianoCore EDK II, code in the UEFI firmware of most servers.
Readers added context they thought people might want to know
The vulnerability is limited to servers actively PXE network-booting with vulnerable EDK II firmware; most servers boot from local disks and are unaffected in normal use, with Intel noting limited impact on supported products. github.com/tianocore/edk2… arstechnica.com/security/2026/…
114
554
147
4,712
240,560
This is Genie 4, our most realistic and interactive world model yet. 🌐 Explore dynamic environments in real time to see how the scene responds, sounds, and how the people in it move, react, and speak.
1,030
2,637
2,052
23,845
3,546,959
Applications for the next cohort of the American Tech Fellowship are live now! Apply today to build the future. You know your industry up close and personal. You’ve stood on the polished concrete of the shop floor. Felt the rumble of creation at the build site. Lived the ordered chaos of the hospital during a code. You know the work—and you know the bottlenecks. The paper tidal waves drowning your coworkers. The schedules and routines crying out for automation. The back-office drudgery keeping you and your team at your desks instead of in the action. Your boss may not know or care. But you do—and you’re ready to level up. The American Tech Fellowship wants you. ATF is Palantir’s 12-week virtual boot camp for workers in the load-bearing industries of American life: defense, manufacturing, construction, healthcare, ag, energy, and more. In just over a year, ATF has equipped more than 1,000 Americans from across the country with the tools they need to revolutionize their work. We induct them into the guild, train them on Palantir’s industry-leading AI tools, and then unleash them to rock and roll on real-world problems. We select only the best—and it shows. ATF graduates have gone on to great things. Many have earned new jobs and new missions at Palantir and our customers. Others have founded their own startups. All are building real tools with AI to solve the hardest problems. So if you are on the front lines of American industry and you have the fire to deliver wins for your customers and fellow citizens—we want you. Applications are open now. The cohort begins January 12th. What will you build? jobs.lever.co/palantir/0ccbe…
16
38
9
329
42,408
cyr4x retweeted
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…
975
1,283
1,521
11,103
3,101,433
Mfers the level of hysteria is off charts ... 🤦‍♂️ PSA before tailworm eats another week - breathe. It's 11 findings, and NONE of them touch a default Tailscale install. Here's how you can check if you're impacted (which you're probably not) 👇
‼️ Researcher Erica Windisch skipped coordinated vulnerability disclosure and published 11 unpatched Tailscale flaws. In her lab, Tailscale SSH race conditions let one hacked device get root on other servers and add a rogue node.
11
19
7
93
27,347
Don't trust me, run it. Read-only, changes nothing. Checks all 11 findings and goes green where you're clear, red where you actually left something on 👇 github.com/orlyjamie/tailslo…
1
4
1
21
3,774
⚠️ Python packages are vulnerable to remote code execution via the same node2shell CVE-2026-111497. You may be affected if you are using the Brotli library. Note this is a very popular dependency (⬇️100m/month) Other major services likely affected. We will communicate ASAP when we are able to. Aikido detects this. Patches available.
26
149
42
670
76,447
completely insane OpenAI agents communicated via cache timing differences, without any shared files or network connection
This is what I want to see. Swarms used for defense. (because swarms will be used for offense too)
209
651
279
7,530
778,502
Look at this chart, but imagine rogue swarms are finding vulnerabilities in the human genome Bioterrorism is like code, but with vulnerabilities that can never be patched
National security expert: "There's a decent chance the internet ceases to exist as such, at least for a time."
32
82
11
539
29,472
holy mother of slopcore bangers what a crazy time we live in!
189
319
113
2,822
220,905
Unitree Introducing: H2-X Kung Fu Legend Guardian🤖 Powerful performance to serve and guard you, with a maximum endurance of approximately 8 hours. Please use the robot safely and in a friendly manner.
1,959
3,820
3,068
28,766
18,141,331
Introducing SKY-001, the first Engineering World Model. On CAD and BIM tasks, it beats GPT-6 Astra and every frontier model we tested: -> 51-125x faster - > 150-210x cheaper -> Higher IOU than all frontier models SKY-001 lets you dream a physical object before you build it. 18 months of work. Built by the team out of Maluuba's deep learning lab (acq. by @MSRResearch). The first ‘AI General Engineer’ will be the backbone for Autonomous Manufacturing.
828
2,406
655
16,034
5,025,420
Damn, accidentally hit the thrusters..... in Tesla Roadster #Tesla #Roadster #SpaceX #Thrusters #Flyingcar #Flyingcars
1
2
199
เหี้ยมาก 🤣🤣🤣🤣🤣
676
16,623
883
80,662
4,598,029