AI for Senior Software Engineers(emadibrahim.com)
emadibrahim.com
AI for Senior Software Engineers
https://www.emadibrahim.com/ai-guide
24 comments
Reminds me of the unfortunate book "Vibe Coding" by Steve Yegge, whom I otherwise enjoy. While it contained okay, if very light on actionable details, overview of the broad ideas behind LLM-assisted coding (how much of it was vibe coding, though?), much of it was co-written through the use of an LLM book editing pipeline, proudly advertised throughout the book. A treatise of otherwise one-tenth of the final length has been blown up into the size of a volume, not unlike a piece of meat is pumped with water to make it appear fattier.
Every time I see a title like this, I ask myself if I'm not being open enough, if my biases are interfering with any potential progress I could be making when it comes to utilising AI. Then I find out that the content is just more slop and it further solidifies my position on all of this. What a waste of energy. It really saddens me.
For me it was the EM dashes in the copy.
that's funny. I read an article about how the use EM prevalance in content as an indicator that it was AI generated. I should tell my agents to stop using them :)
My hat off to the nitpickers and perfectionists
My hat off to the nitpickers and perfectionists
I just came back here to say that.
The author’s LinkedIn profile is also crafted by AI (and the person invented I guess). I’m asking myself if this is a joke, an experiment or a sort of a scam.
I am a real person. Not invented and not a scam, but enhanced for sure :) . I use the tools I have.
I have been in tech for 25+ years and love programming but I have been busy running my company the past few years and too busy to write code and I MISSED IT A LOT, when AI-assisted coding came about, it drastically reduced the friction.
Vibe coding + Years of Experience = Vibe Engineering and it is AMAZING.
I feel like Tony Stark every day. I wake up every day and review code from several agents on several projects and sometimes i start a new agent when i am bed because of a thought/idea I had and when i wake up in the morning the feature i wanted is ready for review and can be live live in under 30 minutes. that is incredible.
I have been in tech for 25+ years and love programming but I have been busy running my company the past few years and too busy to write code and I MISSED IT A LOT, when AI-assisted coding came about, it drastically reduced the friction.
Vibe coding + Years of Experience = Vibe Engineering and it is AMAZING.
I feel like Tony Stark every day. I wake up every day and review code from several agents on several projects and sometimes i start a new agent when i am bed because of a thought/idea I had and when i wake up in the morning the feature i wanted is ready for review and can be live live in under 30 minutes. that is incredible.
Very good then, and sorry for my skepticism! Indeed you’re making good use of LLM on all aspects of your professional life. I’m curious: the same applies to your hobbies and personal life? Any use of LLM that changed also those?
OMG. It has changed everything from “plan me a trip to XYZ for 4 days” to “how do I address my son’s questions about XYZ topic” to “help me create a simple trust for my mom” to spiritual questions to history and philosophy and home automation etc…. I am a woodworking hobbyist and it helped me estimate wood quantities and cut lists and preferred species for different projects and different finishes etc…
The only one it has failed in answering is “how do I make my wife happy” lol
The only one it has failed in answering is “how do I make my wife happy” lol
I am not sure, why invented? Because of a AI enhanced profile picture? If it is AI slop, just flag it. But if it is AI assisted content and the creator is a real person wanting to build up reputation ... then I guess fine by me, if the content is fine. Is it?
(does not look like that to me)
(does not look like that to me)
why would it not be fine if the content is fine but it's fully AI generated? Just curious on why that would not be on with you
As long as the content is fine, all is fine by me.
Yes it is AI generated but I reviewed all the generated code and content and tweaked the design for hours. I am a real person and have been coding since 1995 and have a bachelors from Virginia tech in computer engineering and a minor in math.
AI generation is a tool, and it is the best tool I have ever used in my career. Of course if you don’t know what you are doing, it won’t help you. Just because you have a calculator you don’t become an engineer. Or a paintbrush doesn’t turn you into Picasso.
PS: my wife hates my LinkedIn profile photo too :) she says it looks fake. It’s enhanced for sure.
AI generation is a tool, and it is the best tool I have ever used in my career. Of course if you don’t know what you are doing, it won’t help you. Just because you have a calculator you don’t become an engineer. Or a paintbrush doesn’t turn you into Picasso.
PS: my wife hates my LinkedIn profile photo too :) she says it looks fake. It’s enhanced for sure.
PS: I am the author and owner of that site.
And just to add some more contest, my site used to be a wordpress blog back 15 years ago and I used to block semi-frequently but couldn't keep up with it. The domain has been parked for 10+ years now and i have been wanting to revive it forever but again, too much time than I had to spare. AI tooling has inspired me again and reduced the time significantly. I can now generate content much faster and review it for coherence and accuracy etc...
Lately i have been recording voice memos in the car and when i get home i take the transcript and ask AI to turn it into a blog post. In the past, it just sat on my phone because the effort to edit and research it and format it and type it was too time consuming.
The point is, there is a real person behind this, effort and thought went into it and now i can share my thought without worrying about formatting and grammer and css and blah blah.
And just to add some more contest, my site used to be a wordpress blog back 15 years ago and I used to block semi-frequently but couldn't keep up with it. The domain has been parked for 10+ years now and i have been wanting to revive it forever but again, too much time than I had to spare. AI tooling has inspired me again and reduced the time significantly. I can now generate content much faster and review it for coherence and accuracy etc...
Lately i have been recording voice memos in the car and when i get home i take the transcript and ask AI to turn it into a blog post. In the past, it just sat on my phone because the effort to edit and research it and format it and type it was too time consuming.
The point is, there is a real person behind this, effort and thought went into it and now i can share my thought without worrying about formatting and grammer and css and blah blah.
I feel you brother. Same boat. I've been able to do things I wanted to do for ages and knew how to do, but didn't have the time to type, essentially. I prefer to produce nothing than something half baked ("MVPs"), so I did nothing. Now I can do experiments, throwaway, start from scratch, design, redesign, throw away, and eventually have something I can be a little proud of.
Maybe, just maybe, there will a new era where crappy software (by crappy software I mean almost everything we see around us - open source and closed source) is not the norm anymore - because now a single, smart and knowledgeable enough person can write something like Kubernetes (which is my baseline for OK software) or even Linux in a reasonable amount of time. And maybe we won't be stuck with 40 year old (or more) technologies because "too hard to change". Did you ever try to add bottom padding - literally just bottom padding - to the command prompt in a terminal emulator (in my case, I tried Iterm and Kitty + tmux only)? It's near impossible. That is absurd.
Point is: bad software existed before (most of it), and good software existed before (very little). It will be the same in the future. It's just that now, less people coordinating are needed to right good software.
Maybe, just maybe, there will a new era where crappy software (by crappy software I mean almost everything we see around us - open source and closed source) is not the norm anymore - because now a single, smart and knowledgeable enough person can write something like Kubernetes (which is my baseline for OK software) or even Linux in a reasonable amount of time. And maybe we won't be stuck with 40 year old (or more) technologies because "too hard to change". Did you ever try to add bottom padding - literally just bottom padding - to the command prompt in a terminal emulator (in my case, I tried Iterm and Kitty + tmux only)? It's near impossible. That is absurd.
Point is: bad software existed before (most of it), and good software existed before (very little). It will be the same in the future. It's just that now, less people coordinating are needed to right good software.
> now i can share my thought without worrying about formatting and grammer and css and blah blah
This is exactly what people want you to worry about. Doing otherwise signifies you’re low trust and don’t care about the reader.
The root of the above is: if you were competent, the points you listed above don’t take long and aren’t a meaningful inhibitor to making content. Like coding, AI is fancy autocomplete at this point. If you weren’t competent, you need to get there, otherwise you’re taking a shortcut on your personal brand which is extremely untrustworthy and looks awful. If you care so little about the reader and so much more about producing content, that is not going to come across well.
This is exactly what people want you to worry about. Doing otherwise signifies you’re low trust and don’t care about the reader.
The root of the above is: if you were competent, the points you listed above don’t take long and aren’t a meaningful inhibitor to making content. Like coding, AI is fancy autocomplete at this point. If you weren’t competent, you need to get there, otherwise you’re taking a shortcut on your personal brand which is extremely untrustworthy and looks awful. If you care so little about the reader and so much more about producing content, that is not going to come across well.
I see your point and respect and almost agree. At this point I would rather put something that is 95% good and takes me 30 minutes than spend several hours (that I don’t have) to get it to 100%. The kicker is it will never be 100% because someone will complain or nitpick something.
I appreciate the feedback though. Thanks.
I appreciate the feedback though. Thanks.
> A comprehensive, technical guide to understanding Artificial Intelligence from first principles. Built for experienced engineers who want to deeply understand how AI actually works.
You are overselling this. Being able to read this site and nod along saying "I know this" says enables you to talk to your data scientist or AI engineer.
However, if you don't already know what the jargon means you aren't going to learn it from the site. The prerequisites don't mention probability, but probability isn't introduced anywhere. The sketch about attention in transformers is missing at least one very crucial ingredient. etc.
You are overselling this. Being able to read this site and nod along saying "I know this" says enables you to talk to your data scientist or AI engineer.
However, if you don't already know what the jargon means you aren't going to learn it from the site. The prerequisites don't mention probability, but probability isn't introduced anywhere. The sketch about attention in transformers is missing at least one very crucial ingredient. etc.
You don't even need to read a single word in the article, the second the site opens you know its AI generated, it's that typically crappy white over shadowed rounded card look at that all LLM's seem to love.
That site was generated by AI, wasnt it?
This is getting absurd, seriously. People can choose the tools they use. Whether you're old school or one of the new age people who don't even use punch cards anymore (don't even get me started on people using vi... myself, nothing will ever beat ed).
The comments section of this it's all a version of ad hominem, essentially (whats latin for tool?): "this is using AI, isn't?"
Right now, I'm not interested in the content, so I won't make any comments about that. That's not my point. But the "lol AI????" tone is getting out of hand. Let's judge people by the quality of their content, not by their choice of tools.
The comments section of this it's all a version of ad hominem, essentially (whats latin for tool?): "this is using AI, isn't?"
Right now, I'm not interested in the content, so I won't make any comments about that. That's not my point. But the "lol AI????" tone is getting out of hand. Let's judge people by the quality of their content, not by their choice of tools.
This is AI vomit detected from the first glance. What's the point? LLM access is not exclusive
Surprisingly it was the UI that pinged off the detectors for me.