ok, gpt-4o is incredible. it passes my personal benchmark for laziness.
Before that, only @WhiteRabbitNeos and @AnthropicAI managed to somewhat pass that, with their WhiteRabbitneo33b and Claude-3-opus.
Hella interesting.
Krilled to see what's next 🦐
It’s an honor to announce that Kindo has acquired @WhiteRabbitNeos, the leading creator of open source, offensive cybersecurity AI models. They now have support to further fund & expand the capabilities of its datasets, models, & community. Congrats to @migtissera and team and thank you for delivering an amazing model to the Security community.
Links below: 👇
Important Update: WhiteRabbitNeo & Kindo AI
Today, I’m happy to announce that @KindoAI has acquired @WhiteRabbitNeos.
When I introduced the 1st AI model for WhiteRabbitNeo in December 2023, I mentioned that someone needed to build an AI-focused on offensive cybersecurity and that it would be best to do it in public. This visibility provides a level of transparency and accountability to what is normally a “dark art.”
While executing on that vision we recognized the necessity of either an equity investment or a strategic partnership to effectively scale WhiteRabbitNeo and continue its development. Fortunately, our new alliance with Kindo delivers both.
This acquisition not only secures the essential financial backing, but also merges our mutual commitment and the technical expertise needed to integrate Large Language Models (LLMs) and expand into the information security sector.
The founder of Kindo, Ron (@McclaneDet), who has served as the Chief Information Security Officer (CISO) at 3 unicorns (Riot Games, Bird, and Clover Health), brings decades of invaluable experience. His deep understanding of the security industry's challenges and his proven track record in navigating and addressing these issues make him an indispensable asset to WhiteRabbitNeo and its users.
We’re really excited by our future prospects with Kindo and the ability we’ll have to keep growing WhiteRabbitNeo in the community of forward-thinking cybersecurity developers looking for an edge.
I thank everyone so much for your support, insights, conversations, and words of encouragement. And a special shoutout to all the active participants on our Discord server. WhiteRabbitNeo gets even better from here!
Announcing @whiterabbitneos the creator of the World’s first open source cybersecurity AIs has been acquired by @kindoai. Security AI is now democratized for coaching, secure code analysis, red team exercises, security alert & log explanations, network gear config checking, etc.
Contrary to some media reports, WhiteRabbitNeo is not associated with HackerGPT in any way.
WhiteRabbitNeo is a standalone product, and we have our own fine-tuned LLMs.
We don't have any association with HackerGPT.
At the request of the open source community, we're now releasing a 7B model for offensive and defensive cybersecurity. This can be run locally in most computers with less GPU VRAM.
Access the model here: https://t.co/o7NdmGFp3s
It's been just over two months since we released WhiteRabbitNeo, and we're now at 17K users!
We appreciate all of you! As a gift, use the below coupon code to redeem a 1-month Pro subscription (valid until midnight Sunday).
Coupon Code: FRIENDS
https://t.co/IwJy9Q2WsE
Haha. Tried this for GPT-4, on my personal benchmark. GPT-4 managed to get better through "don't replace code with comments", but not without issues. Apparently, laziness is mostly in the system prompt.
@WhiteRabbitNeos still a bit better.
p.s. also works for custom instructions
Doing another 180 degree turn today.
When I first introduced WhiteRabbitNeo, I said someone has to do this, and it's better done in public.
People have spoken!
WhiteRabbitNeo is back on HuggingFace, with an extended licence and a Terms of Usage clause.
WhiteRabbitNeo is for red, blue and purple teams. Use it responsibly!
Also, consider becoming a Pro subscriber at https://t.co/xDYvX7UE0h