4 ms·

Most of that article is just good career advice in general, so I'll just comment on the part about AI. One major problem I see with the use of AI is that it wi

Most of that article is just good career advice in general, so I'll just comment on the part about AI. One major problem I see with the use of AI is that it will prevent people from building an understanding of <insert problem domain X here>. This will reduce people's ability to drive AI correctly, creating a circular problem.

subhobroto 7mo ago
> One major problem I see with the use of AI is that it will prevent people from building an understanding of <insert problem domain X here>. This will reduce people's ability to drive AI correctly, creating a circular problem. Very much the opposite. LLMs do a fantastic job of increasing accessibility of knowledge. They have wide exposure to content and is incredibly good at spotting patterns and suggesting both well established norms from the current domain and serendipitous cross domain concepts. I feel the concern you share is LLMs expose new frontiers to people who otherwise might not even have imagined that frontier exists and then those people do a lazy or superficial job of it because they lack any internal motivation to do a deep dive on it
- bovermyer 7mo ago
  We must be from _wildly_ different backgrounds.
  - subhobroto 7mo ago
    It's possible we are from very similar backgrounds but bring wildly different perspectives. I'm always suspicious of statements of the form "X will prevent people from Y". The amount of progress humans have made over 2000 years, especially the last 100, is just phenomenal. If anything, there's very little evidence that any X prevents people from Y. X certainly might not require people to Y anymore but I don't question what a motivated person can do just like I don't question what an unmotivated person can refuse to do. It would help if you expanded on "it will prevent people from building an understanding of <insert problem domain X here>", but could one interpretation of it be "AI will prevent people from building an understanding of good software design patterns" because AI allows people to just vibe code and pay no attention to design patterns at all?
    - bovermyer 7mo ago
      Let me start by asking you this: Where does the majority of an individual's general knowledge come from?
      - subhobroto 7mo ago
        their environment and life experiences
        bovermyer 7mo ago
        Alright, so we have our first point of divergence. For me, the majority of my general knowledge comes from reading. Where does the majority of an individual's _understanding_ come from?
        subhobroto 7mo ago
        > Alright, so we have our first point of divergence. For me, the majority of my general knowledge comes from reading Reading is part of your environment - you didn't read your way into a vacuum - the very tools you use to read right now (language, symbols, syntax) are environmental gifts. Something, even if you're consciously unaware of it, made you read those material. The specific script you read is part of your environment (it's unlikely you will know what symbols in Hindi mean if you didn't grow up in that environment or weren't biased towards it). Your physical brain structure depends on the languages you speak and will physically morph as you change those languages. If you have absolutely no clue what "mutton paratha" is, you won't ever start reading about "mutton paratha" right away. (there's another thread we can discuss about human biology, original thought and serendipity but that is better done over email than in HN comments). If you can hear and see, a lot of your general knowledge comes from hearing and seeing too, even if you're consciously unaware of it or don't acknowledge it. If you saw your dad working on cars all your young life, you might not think you have any idea about cars but you might be surprised. > Where does the majority of an individual's _understanding_ come from? Understanding is more complex than knowledge. Understanding and knowledge are at different levels of abstraction but you cannot understand something you have no knowledge of. Thus knowledge is a pre-requisite to understanding but insufficient. With sufficient understanding, you can synthesize knowledge. Nikola Tesla synthesized knowledge about AC Induction motors - a knowledge that's now taught even in high schools from his understanding of EMF. However given the same knowledge of EMF, many people at the times of Tesla didn't have sufficient understanding or motivation or capacity to birth AC Induction motors. You can be extremely knowledgeable but understand absolutely nothing. LLM's are, I argue, excellent examples of this - they have tremendous knowledge but depends on human intervention and discourse to tease it out (as great as SOTA LLMs are, they are not going to randomly design apps without an initial prompt which provides them agency and biases them towards autonomy). (Again, I'm discarding some details that we can discuss over email. For example: I can strongly argue that LLMs have some understanding of what its knowledge is - like the semantic distance between "mutton" and "paratha" and how they can change across various manifolds in latent space) If we agree on this take on understanding, then I argue that the majority of an individual's understanding comes from their consciousness, when they have exercised those neural networks and processed their knowledge to arrive at inferences and re-enforced (belief) or disassociation (misbelief) of assertions
gloxkiqcza 7mo ago
> One major problem I see with the use of AI is that it will prevent people from building an understanding of <insert problem domain X here>. I don’t really think this is a problem. AI is a tool, you still learn while using it. If you actually read, debug and maintain the produced code, which I consider a must for complex production systems, it’s not really that different compared to reading documentation and using Stack Overflow (i.e., coding the way it was done 10 years ago). It’s just much more efficient and it makes problems easier to miss. Standard practices of AI assisted development are slowly forming and I expect them to improve over time.
- subhobroto 7mo ago
  > It’s just much more efficient and it makes problems easier to miss. Standard practices of AI assisted development are slowly forming and I expect them to improve over time Bravo! IMHO, AI just underscores core high quality engineering practices that any high quality engineer has been practicing already. AI is a tool that provides high leverage - if you've been following practices that allow sloppy coding, AI will absolutely amplify it. If anything, I would guess that the AI assisted future will require engineers to think through the problem more upfront and consider edge cases instead of jump and type out the first thing that comes to mind - the AI can spit out code way faster. There's an alternate, vibe coded universe where engineers just spit out slop but as I wrote in another comment here, there are tools to detect that. These are tools that sound "enterprisey" and that's because before AI, no one else had to deal with such scale of code - it's was just far too expensive to read, update and create PRs. Those boundaries are coming down and now almost everyone who can pay for Oxygen tanks have a shot at scaling Mt. Everest.
- nerptastic 7mo ago
  I’ll bite - I’ve been a dev at a new company for about a year and a half. I had mostly done front end work before this, so my SQL knowledge was almost nonexistent. I’m now working in the backend, and SQL is a major requirement. Writing what I would call “normal” queries. I’ve been reaching for AI to handle this, pretty much the whole time - because it’s faster than I am. I am picking up tidbits along the way. So I am learning, but there’s a huge caveat. I notice I’m learning extremely slowly. I can now write a “simple” complexity query by hand with no assistance, and grabbing small chunks of data is getting easier for me. I am “reading, debugging, and maintaining” the queries, but LLMS bring the effort on that task down to pretty much 0. I guarantee if I spent even 1 week just taking an actual SQL class and just… doing the learning, I would be MUCH further along, and wouldn’t need the AI at all. It’s now my “query tool”. Yeah, it’s faster than I am, but I’m reliant on it at this point. I will SLOWLY improve, but I’ll still continue to just use AI for it. All that to say, I don’t know where the future goes - our company doesn’t have time to slow down for me to learn SQL, and the tool does a fine job - it’s been 1.5 years and the world hasn’t ended, I can READ queries rather quickly - but writing them is outsourced to the model. In the past, if a query was written on stack overflow, I would have to modify it (sometimes significantly) to achieve my goal, so maybe the learning was “baked in” to the translation process. Now, the LLM gives me exactly what I need, no extra “reinforcement” work done on my end. I do think these tools can be used for learning, but that effort needs to be dedicated. In many cases I’m sure other juniors are in a similar position. I have a higher output, but I’m not quickly increasing my understanding. There’s no incentive for me to slow down, and my manager would scoff at the idea, really. It’s a tough spot to be in.

subhobroto 7mo ago

> One major problem I see with the use of AI is that it will prevent people from building an understanding of <insert problem domain X here>. This will reduce people's ability to drive AI correctly, creating a circular problem. Very much the opposite. LLMs do a fantastic job of increasing accessibility of knowledge. They have wide exposure to content and is incredibly good at spotting patterns and suggesting both well established norms from the current domain and serendipitous cross domain concepts. I feel the concern you share is LLMs expose new frontiers to people who otherwise might not even have imagined that frontier exists and then those people do a lazy or superficial job of it because they lack any internal motivation to do a deep dive on it

bovermyer 7mo ago
We must be from _wildly_ different backgrounds.
- subhobroto 7mo ago
  It's possible we are from very similar backgrounds but bring wildly different perspectives. I'm always suspicious of statements of the form "X will prevent people from Y". The amount of progress humans have made over 2000 years, especially the last 100, is just phenomenal. If anything, there's very little evidence that any X prevents people from Y. X certainly might not require people to Y anymore but I don't question what a motivated person can do just like I don't question what an unmotivated person can refuse to do. It would help if you expanded on "it will prevent people from building an understanding of <insert problem domain X here>", but could one interpretation of it be "AI will prevent people from building an understanding of good software design patterns" because AI allows people to just vibe code and pay no attention to design patterns at all?
  - bovermyer 7mo ago
    Let me start by asking you this: Where does the majority of an individual's general knowledge come from?
    - subhobroto 7mo ago
      their environment and life experiences
      - bovermyer 7mo ago
        Alright, so we have our first point of divergence. For me, the majority of my general knowledge comes from reading. Where does the majority of an individual's _understanding_ come from?
        subhobroto 7mo ago
        > Alright, so we have our first point of divergence. For me, the majority of my general knowledge comes from reading Reading is part of your environment - you didn't read your way into a vacuum - the very tools you use to read right now (language, symbols, syntax) are environmental gifts. Something, even if you're consciously unaware of it, made you read those material. The specific script you read is part of your environment (it's unlikely you will know what symbols in Hindi mean if you didn't grow up in that environment or weren't biased towards it). Your physical brain structure depends on the languages you speak and will physically morph as you change those languages. If you have absolutely no clue what "mutton paratha" is, you won't ever start reading about "mutton paratha" right away. (there's another thread we can discuss about human biology, original thought and serendipity but that is better done over email than in HN comments). If you can hear and see, a lot of your general knowledge comes from hearing and seeing too, even if you're consciously unaware of it or don't acknowledge it. If you saw your dad working on cars all your young life, you might not think you have any idea about cars but you might be surprised. > Where does the majority of an individual's _understanding_ come from? Understanding is more complex than knowledge. Understanding and knowledge are at different levels of abstraction but you cannot understand something you have no knowledge of. Thus knowledge is a pre-requisite to understanding but insufficient. With sufficient understanding, you can synthesize knowledge. Nikola Tesla synthesized knowledge about AC Induction motors - a knowledge that's now taught even in high schools from his understanding of EMF. However given the same knowledge of EMF, many people at the times of Tesla didn't have sufficient understanding or motivation or capacity to birth AC Induction motors. You can be extremely knowledgeable but understand absolutely nothing. LLM's are, I argue, excellent examples of this - they have tremendous knowledge but depends on human intervention and discourse to tease it out (as great as SOTA LLMs are, they are not going to randomly design apps without an initial prompt which provides them agency and biases them towards autonomy). (Again, I'm discarding some details that we can discuss over email. For example: I can strongly argue that LLMs have some understanding of what its knowledge is - like the semantic distance between "mutton" and "paratha" and how they can change across various manifolds in latent space) If we agree on this take on understanding, then I argue that the majority of an individual's understanding comes from their consciousness, when they have exercised those neural networks and processed their knowledge to arrive at inferences and re-enforced (belief) or disassociation (misbelief) of assertions

gloxkiqcza 7mo ago

> One major problem I see with the use of AI is that it will prevent people from building an understanding of <insert problem domain X here>. I don’t really think this is a problem. AI is a tool, you still learn while using it. If you actually read, debug and maintain the produced code, which I consider a must for complex production systems, it’s not really that different compared to reading documentation and using Stack Overflow (i.e., coding the way it was done 10 years ago). It’s just much more efficient and it makes problems easier to miss. Standard practices of AI assisted development are slowly forming and I expect them to improve over time.

subhobroto 7mo ago
> It’s just much more efficient and it makes problems easier to miss. Standard practices of AI assisted development are slowly forming and I expect them to improve over time Bravo! IMHO, AI just underscores core high quality engineering practices that any high quality engineer has been practicing already. AI is a tool that provides high leverage - if you've been following practices that allow sloppy coding, AI will absolutely amplify it. If anything, I would guess that the AI assisted future will require engineers to think through the problem more upfront and consider edge cases instead of jump and type out the first thing that comes to mind - the AI can spit out code way faster. There's an alternate, vibe coded universe where engineers just spit out slop but as I wrote in another comment here, there are tools to detect that. These are tools that sound "enterprisey" and that's because before AI, no one else had to deal with such scale of code - it's was just far too expensive to read, update and create PRs. Those boundaries are coming down and now almost everyone who can pay for Oxygen tanks have a shot at scaling Mt. Everest.
nerptastic 7mo ago
I’ll bite - I’ve been a dev at a new company for about a year and a half. I had mostly done front end work before this, so my SQL knowledge was almost nonexistent. I’m now working in the backend, and SQL is a major requirement. Writing what I would call “normal” queries. I’ve been reaching for AI to handle this, pretty much the whole time - because it’s faster than I am. I am picking up tidbits along the way. So I am learning, but there’s a huge caveat. I notice I’m learning extremely slowly. I can now write a “simple” complexity query by hand with no assistance, and grabbing small chunks of data is getting easier for me. I am “reading, debugging, and maintaining” the queries, but LLMS bring the effort on that task down to pretty much 0. I guarantee if I spent even 1 week just taking an actual SQL class and just… doing the learning, I would be MUCH further along, and wouldn’t need the AI at all. It’s now my “query tool”. Yeah, it’s faster than I am, but I’m reliant on it at this point. I will SLOWLY improve, but I’ll still continue to just use AI for it. All that to say, I don’t know where the future goes - our company doesn’t have time to slow down for me to learn SQL, and the tool does a fine job - it’s been 1.5 years and the world hasn’t ended, I can READ queries rather quickly - but writing them is outsourced to the model. In the past, if a query was written on stack overflow, I would have to modify it (sometimes significantly) to achieve my goal, so maybe the learning was “baked in” to the translation process. Now, the LLM gives me exactly what I need, no extra “reinforcement” work done on my end. I do think these tools can be used for learning, but that effort needs to be dedicated. In many cases I’m sure other juniors are in a similar position. I have a higher output, but I’m not quickly increasing my understanding. There’s no incentive for me to slow down, and my manager would scoff at the idea, really. It’s a tough spot to be in.