Start by having the agent ask you questions until it has enough information to create a plan.
Use the agent to create the plan.
Follow the plan.
When I started, I had to look at the code pretty frequently. Rather than fix it myself, I spent time thinking about what I could change in my prompts or workflow.
How many hours have you spent writing code? Thousands? Tens of thousands? Were you able to achieve good results in the first hundred hours?
Now, compare it to how much time you've spent working with agents. Did you dedicate considerable time to figuring out how to use them? Do you stop using the agent and do things manually when it isn't going right, or do you spend time figuring out how to get the agent to do it?
You can't really compare those 2. Agents a re non-deterministic. I can tell Clod to go update my unit test coverage and it will choke itself, burn 200k tokens and then loudly proclaim "Great! I've updated unit test coverage".
I'll kill that terminal, open it again and run the exact same command. 30k tokens, actually adds new tests.
It's hard to "learn" when the feedback cycle can take 30 minutes and result in the agent sitting in the corner touching itself and crooning about what a good boy it is. It's hard to _want_ to learn when you can't trust the damn thing with the same prompt twice.
And then all the heuristics you've learnt change under you and you're stuck doing 100-1000 more hours of learning with a drop in quality during that time.
I create a feature branch, do the work and let it commit. I check the code as I go. If I don't like it, then I revert to a previous commit. Other times I write some code that it isn't getting right for whatever reason.
For Workday, use a very simple resume. No columns, no bullet points (use asterisks), no tables.
There's usually an option to upload another file near the end of the form. After it has filled in the fields using your plain resume, delete it and upload the nicer one.
I'm a DevOps/infrastructure person, and I agree completely. This article won't change that.
They've been great for helping me with work-related tasks. It's like having a knowledgeable co-worker with infinite patience, and nothing better to do. Neither the people nor the LLM give back perfect answers every time, but it's usually more than enough to get me to the next step.
That said, having good domain knowledge helps a lot. You make fewer mistakes, and you ask better questions.
When I use LLMs for tasks I don't know much about, it takes me a lot longer than someone who does know. I think a lot of people - not just infrastructure people - are missing out by not learning how to use LLMs effectively.
I switched to Linux earlier this year. I was a little worried about which games wouldn't work, and I didn't need to worry.
The Steam client works just fine, although I do have to "force compatibility" and choose Proton. So far the games have played fine, including big games like Cyberpunk 2077.
For non-Steam, I'm using the Heroic Game Launcher, and it works great. I'm not missing any games, and I certainly don't miss the Windows BS.
Nobody in any kind of business uses @gmail.com. Maybe some Etsy person selling out of their extra bedroom, but that's it.
Domain names and email hosting is too cheap to use gmail for business. Even if it was a legit business, I'd stay away if they couldn't be bothered to do that minimum.
It's not just the use of @gmail.com, it's the fact that it's clearly an autogenerated address with random chars that tripped me.
If it was claiming to be a small business from [email protected], I'd be less suspicious - but a large corporate not using their own domain for mail, the email address itself, and the then switching to a second email address are all giant red flags for me.
It is quite common among US businesses that are not large enough to afford a dedicated IT department and do not want to spend too much on a MSP. Not the kind of company that would be offering this kind of job, for sure.
Huh. I use Rackspace for email, not Hosted Exchange, just email. Haven't had a burp. Been using it for well over a decade with no problem.
I didn't know that the CEO got replaced a couple of months ago. Given the way they're handling this particular problem I may just go looking for someone else, just in case.
Start by having the agent ask you questions until it has enough information to create a plan.
Use the agent to create the plan.
Follow the plan.
When I started, I had to look at the code pretty frequently. Rather than fix it myself, I spent time thinking about what I could change in my prompts or workflow.
reply