We do our meetings on Zoom and there is a Claude plug-in that was set up so that, during the meeting, we can just say, “Hey Claude” and then tell it to do something it will do that thing
Frequently we use it, during meetings, to make cases for jobs we need to do either that day or in the coming week. Claude has access to our Kanban board and so we can, and do, ask it to not only make or edit cases but also to move those cases around the board for us
After the meeting, this might be Zoom itself, we all get an AI document that goes over all of the things that were discussed in that meeting.
Claude-Code is literally an app you can download and it’s soul purpose is to access files on your computer and make changes to them.
Obviously Claude can write code (that’s what “Vibe Coding” is) but Claude-Code doesn’t just write it in the browser and then print it out on screen, it literally writes to files on your computer. You literally give it access to create, delete, update, and edit files on your machine. Because of this, not only can it write code for us, it can also write, and run, our automated tests. And setting this up so that these tests automatically run every hour or to run the relevant tests per PR is a simple task
So yes, Claude not only writes tests for us, it runs the tests. And since Claude can access the files on our computer we can literally tell Claude to do things in our app so, yes, Claude is the first level of testing
It can also read code, so it reads the code pushed and it reviews it. Claude can also approve or deny a PR if it finds something wrong. We still require two humans to review the code, but Claude goes first so I’m not wasting my time reviewing code with obvious problems
Voice-To-Text is not new and you are well aware of the fact that you can chat with AI right? Literally an AI program called “ChatGPT”. So when I talk it gets converted to text and sent to an AI, who responds and Text-To-Speech turns it back into a voice… so I can literally talk to AI
Which mean, yes, I can talk to AI to navigate a phone system or place an order at a fast food place; and I have done literally that
This isn’t even crazy stuff. This is like AI 101 at this point.
Obviously triggered something here. I’m not going to read all this.
I use AI. I’ve built agents from first principles. I’ve tried at these things. I’ve been down these roads.
I know where we stand right now and it’s not the place that delusional tech bros think we are standing.
You seem very invested in the idea that this needs to work out. Like it’s part of your identify now or something. You are defending this as if I attacked you personally.
“I didn’t read what you wrote but I know you’re wrong”
But yeah, you’re totally a super smart AI guy and we should all listen to you. I mean, you’re too lazy to read a few paragraphs on a web forum but you’ve definitely taken the time to study up on all the things AI is currently doing in the business world
It’s not aggressive, If you would read it you would see all I did was explain how it does what you claimed it can’t do
What part am I lying about?
Can AI not write code? Can it not read and write files on you computer? Can it not use Powershell or execute command line prompts? Does Text-To-Speech not exist?
You’ve had Seri for years, you think that’s just going to vanish now?
Have you not used Google or Bing recently? You get an AI overview at the top of almost every search now
You are delusional. It can’t do even half of these things to a level that adds any value to the situation. A correction is coming.
what?
We do our meetings on Zoom and there is a Claude plug-in that was set up so that, during the meeting, we can just say, “Hey Claude” and then tell it to do something it will do that thing
Frequently we use it, during meetings, to make cases for jobs we need to do either that day or in the coming week. Claude has access to our Kanban board and so we can, and do, ask it to not only make or edit cases but also to move those cases around the board for us
After the meeting, this might be Zoom itself, we all get an AI document that goes over all of the things that were discussed in that meeting.
Claude-Code is literally an app you can download and it’s soul purpose is to access files on your computer and make changes to them.
Obviously Claude can write code (that’s what “Vibe Coding” is) but Claude-Code doesn’t just write it in the browser and then print it out on screen, it literally writes to files on your computer. You literally give it access to create, delete, update, and edit files on your machine. Because of this, not only can it write code for us, it can also write, and run, our automated tests. And setting this up so that these tests automatically run every hour or to run the relevant tests per PR is a simple task
So yes, Claude not only writes tests for us, it runs the tests. And since Claude can access the files on our computer we can literally tell Claude to do things in our app so, yes, Claude is the first level of testing
It can also read code, so it reads the code pushed and it reviews it. Claude can also approve or deny a PR if it finds something wrong. We still require two humans to review the code, but Claude goes first so I’m not wasting my time reviewing code with obvious problems
Voice-To-Text is not new and you are well aware of the fact that you can chat with AI right? Literally an AI program called “ChatGPT”. So when I talk it gets converted to text and sent to an AI, who responds and Text-To-Speech turns it back into a voice… so I can literally talk to AI
Which mean, yes, I can talk to AI to navigate a phone system or place an order at a fast food place; and I have done literally that
This isn’t even crazy stuff. This is like AI 101 at this point.
Obviously triggered something here. I’m not going to read all this.
I use AI. I’ve built agents from first principles. I’ve tried at these things. I’ve been down these roads.
I know where we stand right now and it’s not the place that delusional tech bros think we are standing.
You seem very invested in the idea that this needs to work out. Like it’s part of your identify now or something. You are defending this as if I attacked you personally.
Touch grass, talk to humans face to face.
“I didn’t read what you wrote but I know you’re wrong”
But yeah, you’re totally a super smart AI guy and we should all listen to you. I mean, you’re too lazy to read a few paragraphs on a web forum but you’ve definitely taken the time to study up on all the things AI is currently doing in the business world
Seems legit to me
I don’t need to read it because the length of the message tells me everything I need to know.
Being right in this case would prove self evident very quickly.
There is no argument to be had here. If AI is as useful and as inevitable as you say then why even argue for it so aggressively?
When people shout so loudly about how great ai is, its pretty clear what they are really saying.
It’s not aggressive, If you would read it you would see all I did was explain how it does what you claimed it can’t do
What part am I lying about?
Can AI not write code? Can it not read and write files on you computer? Can it not use Powershell or execute command line prompts? Does Text-To-Speech not exist?
You’ve had Seri for years, you think that’s just going to vanish now?
Have you not used Google or Bing recently? You get an AI overview at the top of almost every search now