x
Joined September 2024
Two critical vulnerabilities have been identified in PHP, a widely used programming language that powers websites and web applications. These vulnerabilities (CVE-2026-109618 and CVE-2026-109619) could allow malicious cyber actors to remotely execute code in PHP core, potentially compromising sensitive data and networks. The FBI urges organizations to take the following steps now: ➡️ If an immediate update is not feasible it is recommended that organizations take the system offline until it can be fully mitigated. ➡️ Immediately update affected PHP installations following the guidance on php.net. Inventory where PHP runs in your environment, prioritizing internet-facing systems and vendor products that bundle PHP. ➡️ Monitor web servers for suspicious activity, including PHP processes unexpectedly launching command shells or downloading files. Ensure appropriate process-monitoring capabilities are in place to help identify potential exploitation. ➡️ If you find signs of compromise, isolate affected systems, preserve your web server and PHP logs, and report it to your local FBI field office or ic3.gov. Those reports help us identify the actors involved and warn other organizations before they become victims. For the latest information, follow PHP’s announcement on php.net, CERT-EU Security Advisory 2026-018, and CISA’s Known Exploited Vulnerabilities catalog.
420
5,267
1,236
24,976
3,236,951
cryptopsy retweeted
deepsek is performing quite nicely today
Claude hasn't hallucinated once today
37
58
6
1,510
23,934
cryptopsy retweeted
“Absurd by 3 or 4 orders of magnitude” Dude legit thinks a 2,000 person company should be merging under 30 PRs a day lmao
Mark my words. Anthropic’s codebase is going to start cracking. They have no fucking idea what it does anymore. 30,000 code changes per day is absurd by 3 or 4 orders of magnitude. They’re just churning tokens, hoping it magically becomes intelligent by itself. It’s pathetic.
106
27
3
2,178
158,771
cryptopsy retweeted
For the cost of 8 years of $200 Claude subs, you can force your team to use shitty quantized versions of DeepSeek v4 Flash running at 30tps. Incredible deal!
Local AI is good enough for a startup. A 10-person startup built its own 128GB-VRAM AI server for $18,000. One Threadripper box sitting locally with 🎮 4× Radeon AI PRO R9700 32GB 💾 128GB total VRAM 🧠 Qwen3.8 + DeepSeek V4 Flash 👥 16 concurrent AI sessions And they're actually pushing it. Qwen3.8-Flash-Next with 16 simultaneous sessions: 📚 48K context/user ⚡ 538.5 tps aggregate decode At 128K context per user ⚡ 365 tps aggregate decode The GPUs are even undervolted to under 210W each. Link in ALT
810
252
131
10,805
910,146
I’m on all social platforms. What they’re up to: X: war IG: comparison YouTube: diluting TikTok: rot
388
404
70
9,371
377,497
cryptopsy retweeted
bro it's too late love u tho but get mogged
We have a huge pivot underway here at Epic, moving developers around to retool for maximum AI acceleration. The aim isn't to do the same stuff with fewer people, but to enable everyone here to build MORE and BETTER stuff, and to bring the benefits to all teams using Unreal.
10
4
47
4,270
cryptopsy retweeted
Terrence Tao waiting for society to shutdown AI and die of cancer so he can keep feeling like the smartest person in the world.
85
172
22
2,780
99,896
‼️ BREAKING: Fires broke out after hackers hacked fridges of hundreds of French supermarkets and forced Danfoss controllers to defrost at up to 50°C. Intermarché counts up to 200 stores, Coopérative U fewer than 100. Paris prosecutors opened a probe.
133
582
193
3,373
466,215
cryptopsy retweeted
holy fucking shit. google deepmind reportedly just solved 9 open erdős problems and proved 44 oeis conjectures. two of those problems had remained unsolved for 56 years. every proof was verified in lean, reportedly costing just a few hundred dollars per problem. this is absolutely insane. the pace of progress is getting ridiculous.
27
53
7
568
29,839
cryptopsy retweeted
🚨 Claude Next model 3 hours and 1,135 lines of code Terrain, erosion, lighting, sky, clouds, lens effects, music EVERYTHING was built from scratch by Claude No pre-made assets Just code RIP OpenAI We're not ready for these model
70
96
17
1,768
125,940
Replying to @saintjavelin
This is his reaction
1
2
45
29,972
A tragic day for Elon Musk. Thousands of his most loyal reply guys went offline at once. 💔
1,373
10,436
728
112,536
8,920,167
cryptopsy retweeted
Replying to @thsottiaux
Resets are like ordering a $5 latte at Starbucks. They forgot your order. You complain, they GIVE YOU A FREE DRIP COFFEE & act like they are doing you a favor. OpenAI has no shame. They halved your usage. Give you resets & they think they are Mother Theresa.
4
3
82
16,255
cryptopsy retweeted
Replying to @GuillaumeLample
ML4 was trained on 3,800 NVIDIA Grace Blackwell GPUs in our European datacenters -- our cluster in Bruyères-le-Châtel built using our Series B fundraise. We are investing heavily in infrastructure, with our Series C and D clusters coming online soon. This additional compute will let us train longer, pursue more ambitious post-training, and iterate faster. We expect large and rapid improvements in the weeks and months ahead. 2/n
156
748
96
16,242
725,094
cryptopsy retweeted
Claude Haiku 5.5 scores 1% on MazeBench It is the first model of its class to have a nonzero score, an excellent conclusion to the Claude 5.5 series
23
9
2
561
24,647
cryptopsy retweeted
haiku 5.5 vs grok 4.7 tested both models with same prompt at highest reasoning available > haiku took 27 mins and cost .8$ in claude code > grok took 35 mins and cost 4.1$ in grok build really impressed by what haiku did and literally mogged grok 4.7 which one won here?
haiku 5.5 vs gpt 6 luna haiku is clear winner against luna in this test, but 6.1 sol is still ahead of both model is super fast and cheap. anthropic really cooked harder with this one but it doesn't beat gpt 6.1 sol at all
42
4
5
260
15,454
cryptopsy retweeted
this is what a european datacenter looks like
Replying to @GuillaumeLample
ML4 was trained on 3,800 NVIDIA Grace Blackwell GPUs in our European datacenters -- our cluster in Bruyères-le-Châtel built using our Series B fundraise. We are investing heavily in infrastructure, with our Series C and D clusters coming online soon. This additional compute will let us train longer, pursue more ambitious post-training, and iterate faster. We expect large and rapid improvements in the weeks and months ahead. 2/n
Readers added context they thought people might want to know
The posted photo shows the Breuillet - Bruyères-le-Châtel train station in France, not a datacenter. Mistral AI's cluster is in a separate modern modular facility operated by Eclairion at the location. fr.wikipedia.org/wiki/Gare_de_B… datacenterdynamics.com/en/news/mistra… usinenouvelle.com/article/ce-dat… baxtel.com/data-center/ec…
161
302
52
9,799
448,590
get cucked normies, hermes fcking hates you
27
cryptopsy retweeted
Thank you
16
1
188
7,691