51–60 of 149 posts
Two years, then we'll know if and how this industry has completely been revolutionized.
By then we'd probably have an AGI emulator, emulated through agents.
Alternative, less phishy link: https://github.com/github/gh-aw This is on GitHub's official account. For some reason GitHub is deploying this on GitHub pages without a different domain?
So this being from github.github.io implies it's published by the "github" account on github.
replace github.com/Masterminds/semver/v3 => github.com/Masterminds/semver/v3 v3.4.0
There is just so much going wrong here.
I’m working on an open source project called consensus-tools that sits above systems like this and focuses on that gap. Agents do not just act, they stake on decisions. Multiple agents or agents plus humans evaluate actions independently, and bad decisions have real cost. This reduces guessing, slows risky actions, and forces higher confidence for security sensitive decisions. Execution answers what an agent can do. Consensus answers how sure we are that it should do it.
I find this confusing: I can see the value in having an LLM assist you in developing a CI/CD workflow, but why would you want one involved in any continuous degree with your CI/CD? Perhaps it’s not as bad as that given that there’s a “compilation” phase, but the value add there isn’t super clear either (why would I check in both the markdown and the generated workflow; should I always regenerate from the markdown whe…
But the implementation is comically awful.
Sure, you can "just write natural language" instructions and hope for the best.
But they couldn't fully get away from their old demons and you still have to pay the YAML tax to set the necessary guardrails.
I can't help but laugh at their example: https://github.com/github/gh-aw?tab=readme-ov-file#how-it-wo...
They wrote 16 words in Markdown and... 19 in YAML.
Because you can't trust the agent, you still have to write tons on gibberish YAML.
I'm trying to understand it, but first you give permissions, here they only provide read permissions.
And then give output permissions, which are actually write permissions on a smaller scope than the previous ones.
Obviously they also absolve themselves from anything wrong that could happen by telling users to be careful.
And they also suggest to setup an egress firewall to avoid the agents being too loose: https://github.com/github/gh-aw-firewall
Why setting-up an actual workflow engine on an infra managed by IT with actual security tooling when you can just stick together a few bits of YAML and Markdown on Github, right?
Surely this won't be a security nightmare.
Github should focus on getting their core offerings in shape first. I stopped using GH actions when I ran into this issue: https://github.com/orgs/community/discussions/151956#discuss... That was almost a year ago and to this date I still get updates of people falling into the same issue.
Ah, the critical problem dilemma. Some percentage of free users become paid users, but the free users take up an unreasonable amount of your time/energy/support. The solution seems simple. Buy their product.
That's why it's an issue.
I want to see where we're at in 2 years, because these last couple of months have been pretty chaotic (but in a good sense) in terms of agents doing things with other agents. I think this is the real wake-up-call, that these dumb and error-prone agents can do self-correcting teamwork, which they will hopefully do for us. Two years, then we'll know if and how this industry has completely been revolutionized. By then w…
Especially on the angle of automatic/continuos improvement (https://github.github.io/gh-aw/blog/2026-01-13-meet-the-work...)
Often code is seen as an artifact, that it is valuable by itself. This was an incomplete view before, and it is now a completely wrong view.
What is valuable is how code encode the knowledge of the organization building it.
But what it is even more valuable, is that knowledge itself. Embedded into the people of the organization.
Which is why continuos and automatic improvement of a codebase is so important. We all know that code rot with time/features requests.
But at the same time, abruptly change the whole codebase architecture destroys the mental model of the people in the organization.
What I believe will work, is a slow stream of small improvements - stream that can be digested by the people in the organization.
In this context I find more useful to mix and control deterministic execution with a sprinkle of intelligence on top. So a deterministic system that figure out what is wrong - with whatever definition of wrong that makes sense. And then LLMs to actually fix the problem, when necessary.