Fuck it, still early but here goes ...
We've just released Monty v1 - a Python sandbox that starts in 1 millisecond, not 1.5 seconds.
I just ran 10k sandboxed scripts in 674ms, something that would take a cloud sandbox > 3 hours.
This removes the biggest drawback of letting agents write code. The future is fast.
Even better, it's open source, you can install it from PyPI, npm or Crates now.
Serviced platform coming soon. Please get in touch if you want to be a design partner!
Who should try it?
⚡ if you care about startup time, use Monty
⚡ if you care about long-lived sessions, use Monty - Monty can be dumped and resumed at any external function call
⚡ if you care about accessing functions in the agent/host, use Monty - Monty makes it trivial to expose local functions into the sandbox
⚡ if you care about scale, use Monty - Monty workers use as little as 2MB of memory, meaning you can run thousands of concurrent sandboxes on a single machine
⚡ if you care about security, use Monty - we've run 3 rounds of bounty program and thousands of researchers have tried to break into our sandbox, meaning it should be secure to run untrusted code
Who should avoid it?
🚫 if you like to take a coffee break while waiting for sandboxes to start, DO NOT use Monty
🚫 if you enjoy the challenge of routing API requests from sandboxes through your corporate network to access state in your agent without exposing secrets to the sandbox, DO NOT use Monty
🚫 if your agent really needs to install packages from PyPI, Monty won't help you yet (spoiler: it probably doesn't)
https://t.co/avO8TWY5gJ
Don't fret. They'll just tell us the models will make it better and faster and there's every possible way their stuff just defeats physics and reality.
The paper he is referring to is BitWhisper from 2015. The cpus had to be 0-40 cm apart, both had to be compromised and the bitrate was 1 to 8 bits PER HOUR.
That’s how misaligned AI will kill us?
Maybe we will all die of boredom waiting to decipher what they are doing.
everyone who understands the first thing about computer security or what superintelligence means understands this is possible and all the usual gang of idiots is calling this scifi hype
@PessimistsArc Right. Dario was already claiming that GPT2 was too dangerous to open source back in 2019.
I made fun of them then.
Everyone should make fun of them now.
It's coordinated fabrication being used to sway the public, governmental oversight committees, regulators, and those who don't have the experience that open models are insecure, closed models are more dangerous and far more capable than they currently are, and that a corrupt company should be evaluating all of these incidents.
I'm willing to put my entire career behind it as well.
We've reached the point where, after 25y in cybersecurity, I feel morally obligated to say this for the record:
The narrative being pushed around AI safety, sandbox incidents, and the suggestion that METR be treated as an authority is dangerous, deceptive, and morally corrupt.
Some great ideas here from the cartel:
- you need to give us employee-level access to your entire operation
- if we don't think you're 'safe' enough, sorry we're shutting you down for 'safety'
- China won't comply, but everyone else has to! or no chips!
Brilliant stuff.
Who are we concerned about here? OAI/Ant? The Chinese Labs? Cohere? So far the extent of the risk comes from a) OAI/Ant's models hacking when being asked to hack in poorly secured sandboxes, b) China opensourcing models that increasingly can contribute to nefarious use.
a) doesn't require a 3rd party auditor, O/A need to stop using such weak sandboxes when running such high-risk tasks. b) is a discussion with China, and again is not going to be fixed with a 3rd party auditor.
The 3rd party auditor is a great means of power-projection because you can now arbitrarily raise a barrier to entry by shutting down lower-resourced players, while simultaneously providing zero fixes to the problems at hand, but at least you make people feel good.