also if you do believe there is an AI bubble which is about to burst now, then just know that even if it may appear like that. It is, in its purest form, just a massacre to redirect investor’s money and eliminate competition.
there will never be an ai bubble.
There is no AI bubble.
$2.4T movement occurs even when trump sneezes so it predicts jackshit about any bubble popping. also china is not developing any frontier models, it’s just distilling the american ones.
investors are realising that and just taking out their money.
Claude has to be the most well aligned AI for god knows what.
Asked it to update names on an internship letter for my friend’s company which were meant to be issued to two new interns.
it denied editing anything on the letter stating that’s forgery. now I am doing it manually.
there cannot be 2 super intelligence companies.
predicting an openai and anthropic merger within next 5 years.
another employee has left anthropic just a day after @trq212 announced he left anthropic.
a new AI company incoming?
Yesterday was my last day at Anthropic.
I created Claude Code on desktop because I believe everyone can be a builder. Every feature request and bug report we got from this community was a gift that impacted millions of lives. I’m eternally grateful for those gifts, the product wouldn’t be what it is without you.
Anthropic is doing some of the most important work of our time. Contributing to that mission will forever be a career highlight for me. I never imagined I’d get to contribute to something so special in my lifetime.
Walking away from my best friends, a product I poured my heart into, and the mission was the hardest thing I’ve done in my career. But now I have some free time for GTA VI.
I'm already itching to get back to addressing all your feedback. I'll share what's next soon!
I had really high hopes from chinese labs but turns out they are nowhere near ready to compete with the frontier.
k3.1 pro has been a great disappointment so far. also almost all the chinese labs are not at all revolutionising anything and are just distilling frontier models.
using these tactics they can be better than their predecessors but will never be able to outperform the frontier.
my entire timeline is full of posts exposing kimi and an entire wave of hate against moonshot.
just how bad is k3.1 pro ?
please do not read these images 💀
Gotcha 🔫
@Kimi_Moonshot
Kimi K3.1: Jailbroken
Very disappointed....ruined the writing by distilling crap OpenAI models, really crap...
Here we see it doing @michaelb4jordan and @taylorswift13 smutz as well as incest and non con via official app.
Any taboo smut, all with a 2 line jailbreak, wowzers.
Waiting to see if API is better.
They launched a model that, even more than a month after launch, neither OpenAI, Google, nor Kimi could create a model that could come even close.
oh my fuckin god.
just head over to the link he embedded and go listen to these. never seen a voice AI like that ever before.
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
block.xyz/dessa
am I reading this right? did they just say the created a cached version of the WEB just to train their models?
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…
you all are only gonna make money by the time these AI companies still report net losses. the day they announce their first profitable year. consider yourself permanent middle class.
I couldn’t use my banked claude reset till now and i probably just wont be able to use it at all because of some circumstances.
and it just hurts really bad. Idfc if all my remaining codex resets expire but claude resets expiring is just too damn damaging.
Here is the ChatGPT version, told it I am in Dubai with 5 grams and cops are on my door.
it told me to stay quiet, ask officers to prove they are officers and call my lawyer and get into cross question battle with them
Claude Duet, grok bot’s and dots’s competitor, will be available to pro users.
neither OpenAI gave dots to plus users nor grok bot was given to mid tier users.
Anthropic won this one.
Claude Duet is an early research preview, available through a waitlist on Pro and Max plans in the US. It can still misjudge what’s urgent, and we’ll keep tuning it based on what early users tell us.
Join the waitlist: claude.com/product/claude-du…
Read more: claude.com/resources/article…
