Transcript

00:00:00Hi, I'm Max, maybe you know that, maybe you don't, maybe you saw some videos of me in
00:00:06the past or heard some of my episodes since this is also published as a podcast and this
00:00:11episode here may come as a surprise or maybe not, I don't know.
00:00:17But I have two hearts pounding in my chest when it comes to AI and I have mixed feelings
00:00:29about AI and yeah, I guess many people have that.
00:00:34But if you take a look at the content I publish, at the videos I share, the courses I share,
00:00:41I created courses on ClotCode and Codex and I got a sale right now if you want to get access
00:00:46to all my courses, which include many programming courses too, of course, then you certainly
00:00:52could feel like I'm all-in on AI and AI is everything you should care about as a developer.
00:01:00And yeah, I am pretty much kind of all-in, which I guess is not the definition of all-in
00:01:09now that I think of it.
00:01:10But yeah, I'm pretty bullish on AI and yeah, I think as a developer, you must be able to
00:01:19use it if you want to get a job, for sure.
00:01:23If you want to keep your job, for sure, employers will expect that.
00:01:26And even if you don't care about jobs at all, AI tools like ClotCode can give you that speed
00:01:32boost.
00:01:33They can make you more productive.
00:01:35It certainly feels like that to me at least.
00:01:38At the same time, I'll be very honest, AI has these very depressing aspects and moments and
00:01:46I'm struggling a lot with that and I know many of you do too.
00:01:50You can read it everywhere in comments below the videos on X, lots of blog posts.
00:01:58It's super challenging.
00:02:00And don't get me wrong, I am using tools like ClotCode all the time.
00:02:05We recently relaunched ahead of mine.com, built that a lot with AI, with help of AI, I mean.
00:02:13Looking for build my graphic, maxed, draw, all these projects and bunch of internal tools
00:02:18built with help of AI, steered by me, controlled by me, reviewed by me, but built with help
00:02:25of AI.
00:02:27And then of course there is the entire part where you could use AI to vibe code, utility
00:02:34programs, internal programs.
00:02:36I'm using AI heavily on my machine or the remote servers I'm managing with the Pi coding agent,
00:02:42for example, which under the hood uses my codec subscription, to do all kinds of stuff
00:02:47which I don't necessarily know how to do or I don't have the time or willingness to figure
00:02:52out right now.
00:02:53I mean, for example, just a tiny example, I have a new mouse, wow, big news.
00:03:01And the mouse wheel was inverted, so basically working in the opposite way than I expected
00:03:05it to.
00:03:06And I found no way of changing that on my Mac.
00:03:09I would have to install the software by the manufacturer of the mouse and register there
00:03:16and it didn't feel like that.
00:03:17So I just asked AI, I asked Pi running on my Mac, "Hey, can you figure out the way or write
00:03:23a little program that intercepts the mouse wheel off that mouse and just inverts it?"
00:03:28And yeah, sure enough, it was able to do that.
00:03:31Is that a program I would distribute or sell?
00:03:34No, but it got the job done for me.
00:03:37So AI has these exciting and enabling aspects.
00:03:42It's great for learning, for helping you with learning, for asking follow-up questions.
00:03:48It's also great for not learning anything at all, to be honest, if you just use it to get
00:03:54the answers or the solutions and you don't care about why something works like I did with
00:03:59my mouse program there.
00:04:00So there is always the danger of using it just as a shortcut and getting dumber and dumber.
00:04:06But it has all these aspects and it can make you more productive for sure, feels like that
00:04:11to me.
00:04:12But I'll be very honest, it can also be super depressing.
00:04:17And I've shared that before and that is what I meant.
00:04:19Maybe this video, maybe this episode is surprising.
00:04:23Maybe it's not because I've shared that before.
00:04:26I genuinely liked writing code.
00:04:30I liked learning new technologies, new frameworks, new programming languages.
00:04:37And AI is taking that away from me.
00:04:40And yeah, of course, fundamentals stay important.
00:04:44Understanding the code that gets written stays important.
00:04:47So you still have that learning aspect, but it's not the same as it used to be.
00:04:52And the writing code part is definitely going away for me.
00:04:56It's not 100% away yet, but it's totally different than it was two or three or four years ago,
00:05:03of course.
00:05:05And that is really sad and depressing because you can tell me all day that it was always
00:05:12about solving problems, that it was always about building stuff.
00:05:17Yeah, no, it was always about enjoying what I do.
00:05:22And I genuinely enjoyed the part where I just got into the flow and wrote code and now relax,
00:05:31solved problems.
00:05:32But by coming up with solutions in code, by trying different solutions.
00:05:37And if you're a developer and you started before AI, you know that feeling of working on a problem
00:05:43the whole day and then figuring out the solution the next morning whilst you're in the shower
00:05:48or right when you went to bed and trying it out and making it work.
00:05:54That is super, super exciting or that was super, super exciting.
00:06:00And yeah, I don't have that anymore with AI because sure, you still control and steer AI,
00:06:09but I don't own the code anymore that gets generated.
00:06:14And yeah, you also may get lazy and just accept some solution instead of the best solution
00:06:22or some workaround around the problem instead of a proper fix.
00:06:28And the entire art of crafting code and implementing solutions there and building up a program step
00:06:38by step like this is gone, is gone.
00:06:41And of course you still do that with AI.
00:06:45You build a program step by step by prompting for the different pieces.
00:06:50But my work has shifted from thinking about a problem, writing the code, improving the
00:06:59code and so on.
00:07:01It has shifted from that to writing specs, reviewing specs, having AI generate plans based
00:07:08on specs for the implementation, for parts of the implementation, discussions with AI,
00:07:16correcting or controlling AI, steering AI, reviewing code, drafting tests or telling AI
00:07:25very explicitly which tests, unit tests and so on to write, because by default it will
00:07:30happily write tests that just test the happy path or that are meaningless.
00:07:39Put it in other words, my work has shifted to the parts that were never fun.
00:07:45And that may be very different for other people.
00:07:47And you may not care about all these things.
00:07:49Maybe you're having a good experience with just vibe coding away and not caring about
00:07:54the code.
00:07:55And maybe the future is that you don't need to care about the code anymore because AI models
00:08:00will get better.
00:08:02Well, I'll be very honest, I don't give anything about that.
00:08:05I don't effing care about that at all.
00:08:10For me, the joy was writing the code, working on the program myself, and that is taken away
00:08:17by AI.
00:08:21And yeah, there is a reason why I never wanted to be a manager.
00:08:25I never cared about code reviews and all that stuff.
00:08:28I enjoyed working on my own because I liked writing the code and working in the trenches
00:08:34there.
00:08:35And here we are.
00:08:36Here we are.
00:08:37And don't get me wrong, I want to reemphasize that because I've been very negative over those
00:08:45last minutes because yeah, that is one very strong feeling I have in myself.
00:08:50But as I mentioned initially, and as I mentioned in those other videos and episodes I shared,
00:08:54of course AI has these enabling and exciting aspects.
00:08:58And no matter how I or you feel about it, it is here and it will stay.
00:09:05I don't know how good it will get.
00:09:08I don't know if it will be able to write entire tools end to end in three or five years from
00:09:17now.
00:09:18I don't know that, obviously.
00:09:20Nobody does.
00:09:21Everybody who tells you that he does is just trying to sell you something.
00:09:25I don't know that.
00:09:26So yeah, it is here to stay and you need to be able to work with these tools.
00:09:30You can't just ignore them.
00:09:32At least if it's your job, I think you can't just ignore them.
00:09:36Obviously, nobody's stopping me from still writing code by hand as a hobby.
00:09:41Sure, I can do that just as I could still use a typewriter to write letters.
00:09:47But I don't get a lot of joy out of doing something in a way where I know that it's less efficient.
00:09:54And ultimately, programming software development is my job.
00:09:59I'm a software developer.
00:10:00I'm not doing this just as a hobby.
00:10:02So this isn't a great solution for me.
00:10:06And I'm still figuring out what my solution is.
00:10:10Obviously, it is going with the flow, trying to embrace these tools, learning how to efficiently
00:10:16use them, sharing that knowledge with you in those videos, in those courses.
00:10:22Still teaching fundamentals, which I do and which I will continue to do and where I will
00:10:26continue to create courses on.
00:10:28It will not just all be about AI tools.
00:10:31I try to do all these things.
00:10:35And I hope that is something that will be relevant to you and to the world in general in a year
00:10:42or two.
00:10:43But yeah, AI is that really weird thing, which I sometimes feel like a very small amount of
00:10:52people want it.
00:10:53And yet here we are.
00:10:54And maybe it's like this with every technical revolution, like technological revolution,
00:11:00I mean.
00:11:02And maybe it will be totally different in a few years.
00:11:05I don't know.
00:11:08But on this channel, I want to be honest, obviously, and I am.
00:11:15And I want to share what I'm excited about.
00:11:17I want to share what I think is valuable to you or what I want to get out of my head.
00:11:23And yeah, these two hearts, these two ways of feeling about AI, that is very strongly
00:11:31how I feel.
00:11:32And I think I said that in many of my live streams and videos in the past, but maybe not
00:11:37with as much clarity.
00:11:38So yeah, here we are.
00:11:40And obviously, I want to know what you think and feel.
00:11:45And so yeah, please, please let me know what your thoughts are.
00:11:51And yeah, I'll continue doing what I do, obviously, I'll continue using AI, there is no way around
00:11:57it.
00:11:58I'll continue trying to stay strong in those programming fundamentals and figure out ways
00:12:04of getting more joy out of this new way of building software than I am currently necessarily
00:12:13do, at least not on every day.
00:12:17So yeah, I guess that's it.

Key Takeaway

While AI is an essential evolution for developer productivity and career longevity, it fundamentally alters the creative joy of programming by shifting the role from a builder to a supervisor.

Highlights

Dual perspectives on AI as both an indispensable productivity tool and a source of professional melancholy

The necessity of AI proficiency for modern developers to maintain job security and meet employer expectations

Practical applications of AI ranging from large-scale project builds to solving small personal utility problems

The loss of 'flow state' and the creative joy of manual coding as work shifts toward management-style tasks

The ongoing importance of software fundamentals and manual review despite the rise of automated generation

A commitment to adapting to the new technological landscape while maintaining honesty about its emotional impact

Timeline

Introduction and the Dichotomy of AI

Max introduces his complex, 'two-hearted' perspective on Artificial Intelligence, noting that while he remains bullish on the technology, he harbors mixed feelings. He acknowledges his public persona as an AI advocate who creates courses on tools like Claude Code and Cursor. From a career standpoint, he argues that developers must master these tools to remain competitive in the current job market. Employers now expect a level of speed and productivity that is difficult to achieve without AI assistance. This section establishes the tension between professional necessity and personal sentiment that defines the rest of the discussion.

Practical Successes and Utility of AI Tools

The speaker details how he uses AI for major projects like relaunching websites and building internal tools while maintaining strict human oversight. He shares a specific anecdote about using a coding agent to create a script that inverted his mouse wheel settings on a Mac. This example illustrates how AI can solve 'utility' problems that a developer might not have the time or interest to solve manually. However, he warns that using AI as a constant shortcut can lead to a decline in personal skill and knowledge. He balances the excitement of these enabling tools with the looming danger of intellectual stagnation.

The Loss of Joy and the 'Flow State' in Coding

Max expresses a deep sense of sadness regarding how AI has diminished his genuine enjoyment of writing code and learning new frameworks. He reminisces about the unique satisfaction of solving a difficult problem after sleeping on it or thinking about it in the shower. With AI, that sense of ownership over the solution and the 'art of crafting code' feels largely diminished or gone entirely. He argues that the process was never just about the end product, but about the intrinsic pleasure of the work itself. This section highlights the emotional toll that automation takes on passionate professionals who value the journey over the destination.

The Shift from Builder to Managerial Roles

The discussion pivots to how a developer's daily workflow has shifted from writing and improving code to writing specs and reviewing AI-generated plans. Max explains that his work now consists of steering AI, correcting its output, and ensuring it doesn't write meaningless unit tests. He describes this shift as moving toward the administrative and managerial tasks he intentionally avoided earlier in his career. This transformation is particularly frustrating for those who entered the field to 'work in the trenches' rather than perform code reviews. It marks a fundamental change in the identity of the software developer in the AI era.

Embracing the Future and Maintaining Fundamentals

In the concluding segment, Max emphasizes that despite his negative feelings, AI is here to stay and cannot be ignored by professionals. He compares manual coding in the age of AI to using a typewriter—possible as a hobby, but inefficient for professional work. He commits to continuing his educational work, focusing both on AI tools and the essential programming fundamentals that remain relevant. He invites his audience to share their own feelings about this technological revolution, acknowledging that we are all navigating an uncertain future. Ultimately, he resolves to find new ways to extract joy from this evolved method of building software.

Community Posts

View all posts