NVIDIA unveils NemoClaw

[ggmopa]

It is a hardware agnostic open source enterprise platform that essentially commotises the agent orchestration layer to deploy AI agent workforces as a secure and enterprise ready alternative to OpenAI's OpenClaw.

And it will essentially allow a massive AI workforce to become easy to produce for Fortune 500 companies who will be able to deploy thousands and thousands of agents to run 24/7.

They are already pitching this to all the tech goants. If they nail it, the biggest limiting factor for these conpanies will be hardware. Not software or human workforce.

Read more

 

[Preying Mantis]

And it will essentially allow a massive AI workforce to become easy to produce for Fortune 500 companies who will be able to deploy thousands and thousands of agents to run 24/7.

Yeah cause Fortune 500 companies need more of a reason to think they can replace their workforce.

Not going to be good for the economy in the slightest.

 

[t2van]

Yeah I saw this in my feed but forgot to share.

They quietly launched it this week?, it seems compared to most AI hooha that goes on these days.

It's all free as well isn't it. Locally run, at the moment it's for repetitive tasks it can get smarter while doing them so those tasks become quicker.

The downfall of it is if say you deploy 1000s of AI agents all doing the same thing, making API calls, accessing data, executing code and tasks - who verifies the identity how do they know who done what and then more importantly who takes accountability there's a small element of "trust" missing with it.

From from the data, function and accountability that's the key issue I think.

 

[ggmopa]

Yeah I saw this in my feed but forgot to share.

They quietly launched it this week?, it seems compared to most AI hooha that goes on these days.

It's all free as well isn't it. Locally run, at the moment it's for repetitive tasks it can get smarter while doing them so those tasks become quicker.

The downfall of it is if say you deploy 1000s of AI agents all doing the same thing, making API calls, accessing data, executing code and tasks - who verifies the identity how do they know who done what and then more importantly who takes accountability there's a small element of "trust" missing with it.

From from the data, function and accountability that's the key issue I think.

This is going to be a bit different. Yes it's a bit of a shit show right now with the cluster fuck of security and volatility problems of all the platforms out there. You have to really go out of your way to ensure proper ID verification of agents, using Zero Knowledge proofs network and such things.

But this is going to be different. They will already have multi-layer security and privacy tools built-in that will no doubt be designed integrate seamlessly with NVIDIA's existing AI ecosystem. I bet they will fine tune the models to work specifically with their chips in the future. Still early days but i'd say this is noteworthy; more than just another fancy dashboard - I think it could well become to Agentic AI what AWS became for web and app developers - much better focus on security, privacy, lower need for monitoring but what will make it different is emphasis on shared Infrastructure, rather than small personal setups. Better scalability under one coherent and specialised ecosystem. Just as Amazon did with AWS. Lets wait and see!

 

[Champagne]

Deeper down the rabbit hole it goes.