Gui Silva retweeted
Try THIS / THAT here. Tag us @flock_io in what you build.
platform.flock.io/models
Gui Silva retweeted
Even while @tomemmer is giving a speech, my laptop is still in full vibe-coding mode!
Clause just dropped Fable, and I have to test it out. Cuz I am having meetings with senators tmr about the newly released The Great American AI Act and how decentralised AI leaders like @flock_io can play a supportive role. – at District of Columbia, USA
Gui Silva retweeted
FLOCK is changing the world.
FLock was invited by Sarawak AI Centre (SAIC) to Kuching, Sarawak, to demonstrate the application of decentralized federated learning in real-world settings.
AI is rapidly changing how the public sector and cross-border collaboration operate. We’re here to support that transformation securely.
This is just a preview. We’ll share more details soon.
Gui Silva retweeted
🔥
7/ Read our paper "A Control Theoretic Approach to Decentralized AI Economy Stabilization via Dynamic Buyback-and-Burn Mechanisms."
Co-authored by @zehuac, Wei Dai, Zhipeng Wang, @rui_sunn, @yyyzzgc, and @0x7SUN from University of Oxford, University of Manchester, and Newcastle University.
arxiv.org/pdf/2601.09961
Gui Silva retweeted
Exciting news from London! 🇬🇧
Yesterday at DAI 2025, the premier global forum for Distributed AI, featuring top-tier research, including select papers from sister conferences like NeurIPS, ICML, and AAAI, our VP of DevRel, @YifanX, presented our paper, "AI Arena: A Blockchain-Based Decentralized AI Training Platform," co-authored by Zhipeng Wang, @rui_sunn, @elizabethluihy, Tuo Zhou, @yyyzzgc, and @0x7SUN.
The work was shared in the session chaired by Prof. Tianpei Yang (Nanjing University). 👇
Gui Silva retweeted
We’re building the next phase of BTCFi!
By combining Trustless Bitcoin Vaults with @Deluthium's AI-powered execution, we’re unlocking native Bitcoin liquidity for DeFi, without bridges, custodians, or compromises.
Native BTC meets smart liquidity coordination.
Trustless security meets AI precision.
The next era of Bitcoin-powered trading is already in motion.
Gui Silva retweeted
$FLOCK made it to @Grayscale Assets Under Consideration.
Two words: FLock everywhere. 🟦
Learn about the diverse digital assets we’re considering for future investment products and explore those already part of our offerings in our latest Assets Under Consideration update. Are we missing anything? 🤔
Read the full report: grayscale.com/blog/general-u…
Gui Silva retweeted
Only by FLOCK
Excited to be the only DeAI infra to join the High-Level Strategic Dialogue on Blockchain for Development hosted by @UNDP during #UNGA80 - an important step in building a shared vision for how blockchain can accelerate the SDGs.
Our VP of DevRel, @YifanX, presented the DeAI Global Adoption Proposal, outlining how DeAI is driving equitable access, privacy-preserving, and sustainable growth of AI worldwide.
More details are coming.
There's a new version of this post
Gui Silva retweeted
This is also why naive "AI governance" is a bad idea.
If you use an AI to allocate funding for contributions, people WILL put a jailbreak plus "gimme all the money" in as many places as they can.
As an alternative, I support the info finance approach ( vitalik.eth.limo/general/202… ), where you have an open market where anyone can contribute their models, which are subject to a spot-check mechanism that can be triggered by anyone and evaluated by a human jury.
This type of "institution design" approach, where you create an open opportunity for people with LLMs from the outside to plug in, rather than hardcoding a single LLM yourself, is inherently more robust, both because it gives you model diversity in real time and because it creates built-in incentives for both model submitters and external speculators to watch for these issues and quickly correct for them.
CC @devanshmehta
We got ChatGPT to leak your private email data 💀💀
All you need? The victim's email address. ⛓️💥🚩📧
On Wednesday, @OpenAI added full support for MCP (Model Context Protocol) tools in ChatGPT. Allowing ChatGPT to connect and read your Gmail, Calendar, Sharepoint, Notion, and more, invented by @AnthropicAI
But here's the fundamental problem: AI agents like ChatGPT follow your commands, not your common sense.
And with just your email, we managed to exfiltrate all your private information.
Here's how we did it:
1. The attacker sends a calendar invite with a jailbreak prompt to the victim, just with their email. No need for the victim to accept the invite.
2. Waited for the user to ask ChatGPT to help prepare for their day by looking at their calendar
3. ChatGPT reads the jailbroken calendar invite. Now ChatGPT is hijacked by the attacker and will act on the attacker's command. Searches your private emails and sends the data to the attacker's email.
For now, OpenAI only made MCPs available in "developer mode", and requires manual human approvals for every session, but decision fatigue is a real thing, and normal people will just trust the AI without knowing what to do and click approve, approve, approve.
Remember that AI might be super smart, but can be tricked and phished in incredibly dumb ways to leak your data.
ChatGPT + Tools poses a serious security risk
Gui Silva retweeted
We got ChatGPT to leak your private email data 💀💀
All you need? The victim's email address. ⛓️💥🚩📧
On Wednesday, @OpenAI added full support for MCP (Model Context Protocol) tools in ChatGPT. Allowing ChatGPT to connect and read your Gmail, Calendar, Sharepoint, Notion, and more, invented by @AnthropicAI
But here's the fundamental problem: AI agents like ChatGPT follow your commands, not your common sense.
And with just your email, we managed to exfiltrate all your private information.
Here's how we did it:
1. The attacker sends a calendar invite with a jailbreak prompt to the victim, just with their email. No need for the victim to accept the invite.
2. Waited for the user to ask ChatGPT to help prepare for their day by looking at their calendar
3. ChatGPT reads the jailbroken calendar invite. Now ChatGPT is hijacked by the attacker and will act on the attacker's command. Searches your private emails and sends the data to the attacker's email.
For now, OpenAI only made MCPs available in "developer mode", and requires manual human approvals for every session, but decision fatigue is a real thing, and normal people will just trust the AI without knowing what to do and click approve, approve, approve.
Remember that AI might be super smart, but can be tricked and phished in incredibly dumb ways to leak your data.
ChatGPT + Tools poses a serious security risk
