Making agentic workflows hallucination and poison free powered by our verifiable databases and cryptography tools.provably.ai SwitzerlandJoined April 2024
anthropic.com/research/multi…
This is fantastic read.
1. agents struggle to coordinate with each other especially for common resources.
2. they don't know how to detect when they are being lied to.
And as usual Anthropic wants to solve it through more learning..
SourceryKit detected 100% of the covered deterministic grounding errors in this study.
Next, we will test how much accuracy agents can recover through answer healing and targeted retries.
Read the study:
provably.ai/blogs/The-Agen…
We believe agents should capture evidence natively as they work.
SourceryKit captures tool calls in a state of the art verifiable database, allowing agents to evaluate themselves and downstream agents to independently verify their answers.
We spent 9 weeks testing agent reliability during software tool use.
Nearly 60% of claims accepted as correct could not be backed by evidence from recorded tool calls.
The answer looked right. The evidence wasn't. 🧵
AI is going collaborative.
In the next 6 months all we are gonna talk about is verifiability (for safety, security and accuracy), protocols and crazy new use cases.
The main thing in AI 2.0 - which Buzz is triggering with collaborative AI - is verifiability.
Bolting on & releasing shortly.
Then whenever another AI does a task it, it can prove it really did what claimed it did.
It saves you needing access to the “actual” Terminal / end point.
SourceryKit is available now.
It is a Python SDK for runtime verification of agent tool calls.
You can verify the tool response while the workflow is still running, before the error spreads.
Get started in minutes at provably.ai with
pip install sourcerykit
Check our launch blog: provably.ai/blogs/Runtime-…
Our view:
The future is agents that can verify their own work, catch errors, and heal while the workflow is still running.
As agents get smarter, this feedback loop will help them adapt and continuously learn from new workloads, powered by verifiable data from each other.
It is time to take the training wheels off.
SourceryKit in early adoption:
- detects 100% of agent tool-call errors.
- improves accuracy by up to 50% in some models.
- works across single-agent and multi-agent workflows.
This is why we are building runtime verification for production agents.
All that scary marketing has blown up in their faces?
- this means frontier models in rest of the world should get access to customers and data they wouldn’t normally get.
- the lead they built could turn out to be temporary.
Open source gets 80% of the tokens but only 20% of the revenue. OS struggles to access new data and improve as fast as frontier and closed source.
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of
Personal update:
We found a way for agents to deliver more accurate and reliable results - at runtime.
Here’s a 35 second sneak peek of an agent answering a user question, fetching data from backend systems, and verifying its own response at runtime.
2K Followers 3K FollowingMathematician, working on @zcash.
Interested in applied mathematics. Cryptography for a living.
Physics and abstract algebra in my dreams.
@leak_ix
430 Followers 26 FollowingNazare Ventures is a venture capital firm focused on early stage investments in AI infrastructure. Founded by @deseventral. follow the robot wave 🤖🌊
4K Followers 2K FollowingAssociate Professor of Finance, @UWMadison. I study safe assets, banking , central bank policy- tools to prevent financial crises and clean up after them.
454 Followers 645 FollowingMathematician & cryptography researcher @icme_labs. PhD @ImperialCollege. Summering @a16z. I care about privacy and human flourishing. @zcash is hope
2K Followers 214 FollowingWe advance privacy & cryptography by connecting you to the most promising protocols.
Build squads. Pursue freedom. Become Sovereign.
90K Followers 1K FollowingFounder @AvailProject. Previously founder @0xPolygon. Building ShieldTX for shielded Hyperliquid perps - check out https://t.co/XPeVqncjVc to see who’s copying you
1K Followers 1K FollowingSelf-sovereign but never self made. Learning and building daily.
Developing with a dedication to personal agency.
Founder - AmachHealth
10K Followers 4K FollowingThe PALM Economy | Powered by the $PALM Community!
A decentralized economy for RWAs
Fueled by $PALM on Cardano | $RSPALM on BNB| PALM DAO | On-chain commodities
4K Followers 2K FollowingAssociate Professor of Finance, @UWMadison. I study safe assets, banking , central bank policy- tools to prevent financial crises and clean up after them.
2K Followers 722 Following'picious until proven otherwise @zksecurityXYZ. Research and cryptography. Recurring co-host on @zeroknowledgefm. Also a troubadour and chess enthusiast
19K Followers 30 FollowingPrivately fund your wallet. $500M+ processed. 140K+ users. 20 audits. Backed by @alliance. No CA. Live on Solana, Robinhood, Ethereum, Base, BNB, Hyperliquid.
1.8M Followers 178 FollowingNobel Laureate. Co-Founder & Chair @GoogleDeepMind; Chief Scientist of Alphabet. Founder & CEO @IsomorphicLabs. Building the future...
8K Followers 94 Following✦ Creating world-class saas product videos
✦ @beehiiv and 150+ (early to late stage) startups
✦ Work with me → https://t.co/r4ynMoqVbx
5K Followers 1K FollowingPassionate about crypto, distributed systems, cryptocurrencies, and making computation & communication private.
Tweets will be technical & only my own opinion
630 Followers 238 FollowingAudio technology | Your all-in-one platform that allows you to create, publish, & monetize audio content. Podcasts, audiobooks, talk shows, and more 🎧🎙
4K Followers 2K FollowingBuilding the future of storytelling @JamitHQ | prev @Roku, @NewsCorp | Proud @beondeck Fellow. A Non-Traditional Engineer with a Creative Mindset
29K Followers 76 FollowingStories from everywhere, brought to life for everyone. Discover immersive audio stories personalized to what you love. Create, listen, and enter new worlds.