itigges22

The observability gap is absolutely something that I have faced. I've resorted to --verbose and reading terminal scrollback to understand why Claude made certain decisions. The graph view sounds like the right abstraction - tool calls form a natural DAG. Does it capture the reasoning between tool calls, or just the calls themselves? The reasoning chain is usually where I lose my thread!

In my opinion- we need to successfully integrate generative AI into our workflows, but we need to do it appropriately and exercise caution. Ex. If you are a senior developer utilizing AI as a tool then I see using Generative AI as a competitive advantage. If you are a junior developer claiming to be a senior developer that is not okay.

All in All, just be honest with yourself, and be honest to others on where your skills lie. AI can be a great tool, but it should not replace you.

In my opinion they cannot confidently rely on it- I think its great to get a product out there, but I think we can draw parallels to things like Dropshipping- where the "Get quick rich" thought process is a major failure. Its better to think about it as that you still need a large amount of energy, time, or capital to achieve a certain level of success. (INCLUDING production coding projects that req security and strong engineering practices). We can use Spec Driven AI Assisted coding to add a level of rigor to "vide coding", but I still don't think models are where they need to be in order to get it perfect. We will almost always have issues.

Lots of security vulnerabilities, but it was vibe coded. As much as I am an Anthropic Fanboy- I hope OpenAI can add some level of security to OpenClaw and they can get some good contributions. Although, these issues may be deep to its core.

I very much understand the pain of a broken homelab/ proxmox setup! Has made me want to do something similar for a while! I have dealt with old RAM crashing, PSUs going out, HDD's breaking, Proxmox Kernel Panic (idk how, but it happened for me). Having something that allows me to redeploy with a single command is amazing.

Does this capture animations as well? In threeJS or animation heavy sites AI Struggles to develop it effectively! Have not seen a good solution for this yet.

For security vunerability testing on websites I have been making for clients- I almost always hire a senior developer to look over the work and or tests that were created. AI can pass a test, and it can make something that passes a test, but there almost ALWAYS are problems that the senior dev finds with the tests, or with the code that was being tested. Sometimes AI will adjust the code entirely to pass the test or adjust the test to pass failing code.

Another counter-measure I have is to simply lock code before testing. Look over test files, and ensure its not following the happy path.

I just saw a post from Yann mentioning that AMI Labs is hiring too!