AI Shepherd | https://xitter.cf/t.co/ob5V88TaRc / https://xitter.cf/t.co/ITEBAPNLX6 / https://xitter.cf/t.co/VQ0S2Xfp2f / https://xitter.cf/t.co/4UUmU3WrFp / https://xitter.cf/t.co/tjp9KDBK0T / https://xitter.cf/t.co/Bz7yjs4i2P

Joined March 2025
BIG AGENT is a massive departures board for your coding agents 👀 Now with faster monitoring, bigger text, better colours and a new character icon. Mac release out now 🍎 Unsigned for now—build it yourself or follow the README’s install instructions.
1
1
421
IPO CLAXON 🚨
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.
8
If the massive gulf between what people think AI is capable of and what AI is actually capable of had a brand name, it would be Gemini. Google is single handedly making people think AI is fucking trash.
Replying to @theo
I've said this a few times but I really want to emphasize it: This isn't the usual "gemini is silly and 5-10% behind the other labs". Gemini is a straight up liability in your stack. Any engineer who thinks it's usable should not be taken seriously.
29
if it's written in German it is not the best anything.
This is the best and tightest written show, and nothing will ever top it. I don't want to hear any arguments.
13
Looking at the sheer number of CVEs emerging, I think we're heading for a mass rebuilding of web estates across thousands of orgs. Simpler architecture. Less code. Lighter products. Far fewer dependencies. Rebuilt in a hurry, out of necessity. The Great Simplification.
AAAAWWOOOOOOOGAAAAAAAAA A high severity vulnerability in "the most widely deployed DNS software on the internet" just dropped kb.isc.org/docs/cve-2026-943…
25
*days without a security incident* 😭
introducing pwn.watch 👀 one place for following all the live security incidents it serves rss + llms.txt so u can paste to your clankers tell them to watch it on a schedule and sleep at night 😴 100% free but kitze.io/sponsors will appear on it ❤️
20
are these exploits all coming from people using currently available models? or is this coming from next gen models that are in preview release?
🚨 Nightmare day for cybersecurity Multiple major security incidents, active exploits, supply chain compromises... If you run production systems, start patching NOW. PHP: Two critical unauthenticated RCEs. Attackers already exploiting them. Patch NOW LiteLLM: v1.107.0 on PyPI and Docker is BACKDOORED. Steals cloud, SSH, and AI credentials. Pin 1.106.2. Rotate secrets. eMudhra: A certificate authority's keys were crackable. Update your browsers, systems, and trust stores. Node.js / Next.js / React: Security releases for Brotli RCE, TLS bypasses, and TWO critical Next.js RCEs. TiDB Cloud: PingCAP breached through stolen credentials. Customer information and other sensitive data exfiltrated. Read the attacker's release, fun stuff. Plex2Shell: Weaponized within HOURS of the PoC. Attackers already deploying cryptominers and proxy backdoors. BeyondTrust / Ivanti: Actively exploited zero-days added to CISA's KEV catalog. BeyondTrust attacks linked to a China-nexus campaign. And that's not even the full list. Patch. Audit. Rotate. NOW.
15
Anyone else's codex suddenly transcribing total jibbersh from live voice? It's just getting everything I am saying completely wrong. Was fine yesterday.
14
Sounds like a modern dystopian version of Microsofts employee stack ranking, which was already pretty dystopian.
One-size-fits-all blanket usage caps on AI spend don’t make sense. On the other hand, yolo tokenmaxxing doesn’t scale either. We’ve come up with a better approach to cost control centered around efficiency. More efficient employees can have limits raised automatically (with a final upper bound still), while other spend requests go thru an admin review process.
42
Even after the chat has the input it needs either directly or as a message from another chat, THIS CHIP WILL NOT GO AWAY. So you constantly have to check if it actually needs input. Clearly not a single @OpenAI employee is using codex any more. It's full of obvious UX bugs.
1
101
Imagine being so unimaginative about AI that you make the mass unemployment everyone feared a reality. The US is using AI to do more and hire more people. Britain is using it to sack people instead. What a damning indictment of Britain.
Key point made in this article about needing to know where the banked time savings go. That's a problem I see arise repeatedly. Once AI takes over certain tasks and time-consuming processes, they're not sure where to spend their time.
1
1
67
openAI is solving the mysteries of mathematics. meanwhile, Google:
Google DeepMind says four Gemini 4 Argon agents beat Minecraft without any human help in 27 hours and 14 minutes, costing $1,620 at Argon's planned 2027 price (@maba_xr / The Decoder) (Visit Techmeme dot com for the link and full context!)
20
Woah, I just realised I can have my local codex message my remote codex. 🤯 Just discovered that completely accidentally when my local agent said it could see my remote agent was running a related task and asked could it message the remote agent to coordinate. YES, you may!
2
1
2
50
Bad Gateway Co retweeted
🤣🤣🤣
532
8,912
586
79,200
2,294,141
The trick to staying employable in tech is simply learning the new tech. There's no magic bullet to stay relevant, you simply need to have a voracious interest in whatever is new and be deeply invested in getting good at whatever it is. Just learn the new thing. It's not hard.
This hacker news thread is absolutely heart-breaking. Whoever thinks “humans will find a way” ought I read it.
1
41
"The FBI has since arrested Abdulrahman Alfozan, a 35-year-old Saudi national on OpenAI’s product team" 🫨
BREAKING: OpenAI says it discovered a sophisticated recording device hidden in an executive conference room at its San Francisco offices and used its own AI models to trace it to an employee who had planted it days earlier. The FBI has since arrested Abdulrahman Alfozan, a 35-year-old Saudi national on OpenAI’s product team, who allegedly installed the device to capture sensitive conversations about the company’s technology. OpenAI says the employee has been fired, that it has no indication model weights, source code, or customer data were accessed, and that the FBI’s investigation is ongoing.
84
We've crossed the Slopicon. No going back now.
Thank you Opus 5.5 🫡 Kinda wild that I vibe coded a slop port of the LANGUAGE ITSELF and got a 3x reduction in typechecking
18
meh, did this months ago. opens/edits word docs right in the browser. freedesk.app
Remember the team building free, open-source alternatives to Adobe apps like Photoshop? Well, they’re taking on Microsoft Word too. WordCraft rebuilds the familiar Word workflow in Rust. Free, open source, and already available on GitHub, though still in alpha. All software will become free. I love it.
1
25
Bad Gateway Co retweeted
Some rumors I've heard. I'll start with Google, then Anthropic and OpenAI. Google: Insiders say Google has either cracked RSI (recursive self-improvement) or is very close. Their pace has picked up fast in recent months. They went from GPT-5 level to Opus 5.5 level in months, and speculation is that RSI made that possible. Anthropic: About a week ago, Anthropic started a new pretraining run. It is not Fable 5.5 but what would become Fable 6, or something bigger, to catch up to OpenAI's Bel. OpenAI: Bel (it may have another codename) is the model behind the math results OpenAI published a few days ago, and RSI apparently helped there too. From what I'm hearing, OpenAI, Anthropic, and Google are all very close to cracking RSI or already have. It will get scary and exciting at the same time.
427
840
244
16,320
2,299,035
great. a new spam vector.
Replying to @nullswan
@bot send him an email of a goose
19