I thought for a while there will be this massive standardized schema connecting all World APIs into a single traversable object. Allowing you to easily connect anything.

In this case LLM would have to write a bunch of stuff from scratch though and might call APIs wrongly.

You are not really defining the runtime?

Is whatever what happens between e.g. HTTP Request and Input and Output not runtime then?

1. HTTP Input

2. While (true) runAgent() <- is that not runtime?

3. HTTP Output

Additionally Claude could be triggering itself with custom prompts etc to use instances of it concurrently in parallel.

Or are you saying that the only rule is that Agent is being ran in a loop?

But the whole discussion is about how AI Agent is different from a Workflow?

The point is that workflow is that LLM is triggered in a loop ?

That's just you hedging, they don't really need to know that. As long as if you are hedging accurately in the big picture, that's all that matters. They need estimates to be able to make decisions on what should be done and what not.

You could tell them that 25% chance it's going to take 2 hours or less, 50% chance it's going to take 4 hours or less, 75% chance it's going to take 8 hours or less, 99% it's going to take 16 hours or less, to be accurate, but communication wise you'll win out if you just call items like those 10 hours or similar intuitively. Intuitively you feel that 10 hours seems safe with those probabilities (which are intuitive experience based too). So you probably would say 10 hours, unless something really unexpected (the 1%) happens.

Btw in reality with above probabilities the actual average would be 5h - 6h with 1% tasks potentially failing, but even your intuitive probability estimations could be off so you likely want to say 10h.

But anyhow that's why story points are mostly used as well, because if you say hours they will naturally think it's more fixed estimation. Hours would be fine if everyone understood naturally that it implies a certain statistical average of time + reasonable buffer it would take over a large amount of similar tasks.

Define runtime then.

> If your system receives 1000 requests per second, does it keep writing code while processing every request, on per request basis? I hope you understand what run time means.

With enough scale it could, however it really depends on the use case, right? If we are considering Claude Code for instance, it probably receives more than 1000+ requests per second and in many of those cases it is probably writing code or writing tool calls etc.

Or take Perplexity for example. If you ask it to calculate a large number, it will use Python to do that.

If I ask Perplexity to simulate investment for 100 years, 4% return, putting aside $50 each month, it will use Python to write code, calculate that and then when I ask it to give me a chart it will also use python to create the image.