Conf42 Large Language Models (LLMs) 2025 - Online

- premiere 5PM GMT

Managing the Gap AI Brings: Keeping Developers Sharp and Engaged

Video size:

Abstract

The introduction of advanced AI-assisted coding tools, namely GitHub Copilot, ChatGPT, and Code Llama, has changed the software development process considerably. While these tools enhance productivity and lessen cognitive load, they pose a critical challenge of skill atrophy. The more AI does the coding, the more programmers stand to lose their ability to tackle complex issues, optimize systems, and comprehend system design on a holistic level. This talk focuses on how businesses and developers can deal with the “AI gap”—the disparity between human-created value and the value created with the assistance of AI. At the end of the session, participants will have in hand a plan of action that ensures the use of AI enabled coding software tools without diminishing or losing the abilities of software development. The intended audience is engineering managers, AI scientists, and computer programmers who seek to use AI but want to avoid the absence of the reasoning faculties and imagination necessary for original work.

Summary

Transcript

This transcript was autogenerated. To make changes, submit a PR.
Hello everyone. I am excited to be here today at 42 to talk about something that affects all of us, how AI is changing our work and skills. AI coding assistant, Chachi, pt, GitHub copilot, cursor winder. They have made our development faster and easier, but they bring real challenges. So we need to ensure that as AI takes over repetitive tasks, we don't lose our ability to think. Critically or creatively and to be en to ensure that if we are the ones that are solving the problem. So that's exactly what we are going to explore today. So let's get started. AI makes coding easier, let's be honest, but it can also make us mentally lazy if you are not really careful about it. So there is something called metacognitive laziness. It happens when we rely too much on something, in this case, ai, and we stop thinking critically and engaging with our work. So this is actually an issue I. And, researcher learned about, something called ai, chatbot induced cognitive therapy or short for ai, C-C-I-C-A. This is where developers, lose their critical thinking skills over the time. So it's not just on a day, but over the time. Just like muscle, we can without exercise our coding and problem solving skills deteriorate when AI does all the work for us. Another big problem in addition to laziness would be skill degradation with AI providing instant answers would skip. The why and the how of coding, which is very important in learning process. beginners in particular would risk developing a shallow grasp of programming concept. Debugging is another issue. AI often finds and fixes the bugs, especially with the new tools such as windsurf and cursor. So this means developers do not get their hands on practice like they used to, and, they wouldn't have the tendency to improve themselves. So over the time, I believe our mental muscle memory would be getting weaker and weaker, and it would make really hard for us to solve complex problems independently. So a real world studies are already showing skill degradation. this is due to the AI reliance. So think of yourself, how often. Do you use chat g pt? three years ago we're using any tool, AI tool, maybe other than Grammarly to check your grammar or something like that. So as a new modern individuals, we are getting more reliant on ai. So actually a study found. That nearly 69%, students reported laziness and, 27 lost their decision making skills because of ai. So experts have drawn. Pilots to aviation because when pilots rely too much on, autopilot, their ability to fly manually, fly a airplane or any sort of aircraft such as, helicopter, it would decline significantly. So the same could happen to, developers and engineers. If we led AI to thinking for us, and we would be just, Additional tool for the AI to handle the system. So how do we combat this? One solution is to build a culture of, continuous learning, which is very important on individual level. many companies provide e-learning platforms. And sponsors professional certifications. And another one would be knowledge sharing, like tech talks, forums, would help developers keep learning from each other. And, another one would be like, something like. hacker rank, which is a platform or lead code, which would allow our developers to practice real world coding challenges and stay sharp. this is just for developers, right? But think of AI engineers. We don't have to many provide coding. We can provide different questions where they would actually go and, practice or do their best or challenge themselves. In this case, hacker rank and lead code would be suitable, for example, developers. another effective strategy is focusing on hands-on projects that would push developers to think critically. passion projects, r and d prototypes and internal coding competitions. These would keep developers actively engaged. So hackathons, are great way to force developers and not only developers who are going and attending as, a team, but also as a judge. It would, encourage them to solve unique, challenges or observe people who are, solving unique challenges and judge them and learn from their, practices. And, someone who's joining, hackathons. I believe the individual would be under the pressure and they would practice working on under the pressure. So it is very valuable. The key takeaway is that we are learning by doing, right? So it is a best way to maintain, unless, in the future we're gonna have chip in our head, we are gonna just install the information. But at this stage, learning by doing, is the best, follow of us. Now one of the best ways to keep developers sharp is, mentorship and collaborative, work. It can be coding, it can be brainstorming. So when experienced developer guide a junior through problem solving, both part would be learning. a way, person who is, for example, a senior developer, they would, revisit their knowledge and at the same time, a junior would be learning from the, team member and try to understand what are the best practices in terms of. Coding or programming or in, in terms of architectural decisions. For example, in pair programming, one person would question the AI generate code and that would lead to a deeper discussion in team. Another one would be, code reviews. If code reviews, require developers, either if the individual is a junior person, mid-level, or senior, it would, give them a opportunity to explain their understanding from generated code. And obviously as a business owner or who, who is less technical, they would feel a more confident on, Written code where we literally used, AI as a co-pilot rather than co-pilot. So it is very important to have code reviews, pair programming, collaborative communication and practices, as a team. So the key is to use AI responsibly without becoming dependent. I would say try solving problem on your own first before using ai. all this would ensure you to engage in what would be the potential solution. And then re reflective practices like, how would solve without ai, would help and may maintaining your critical thinking skills. Also, AI should enhance, right? It shouldn't replace you You do not underestimate human intelligence, although it works completely different in comparison to AI or artificial intelligence. But it is there for you to support you, not to replace you. That is why, make sure that you use it as a co-pilot, you being pilot at the same time. One of the most effective ways to keep developers, engaged and sharp is by exposing us developers to different projects and ideas. So it doesn't have to be project at the same time. we have different problems for same domain, and we've got same problem in different domains, but applied differently. So project rotation. Would help developers work on different code bases and, it would, help problem solve with the problem solving skills and preventing them from, other specializations, which is, very important, especially at this time and age. tech workshops, hack days, dojo, sim, would simulate continuous improvement. and they would give the opportunity to have hands-on collaboration with others. Systems like, by Spotify and ais, they created the communities where developers practice skills together and share best practices. So the key is to encourage peer learning, ensuring that, developers remain adequate and capable of, first of all, understanding, then solving the new problems. AI has significantly boosted, developed productivity and the big deal. We all know that, but especially when it comes to, coding tasks or coding related tasks, it really helped us like. Documentation, writing and refactoring. a McKinsey study found that a developers complete programming task up to two x faster than using AI coding assistance, documentation writing, which is related to coding but not coding itself. a factoring of documentations. They were improved by 50 to 67% with, AI assistance. So it really reduced the time spent on low value. Tasks in, stack Order flows 2024. Developer survey 81 personal developers cited increased productivity. I, myself, personally, I think it's great. It helps with the, writing and understanding of the code, and you are becoming fast. But while AI speeds up the development process, it is crucial that we use it wisely to ensure that we are still building and refining, our problem solving skills, which is very important in the long run. AI is powerful and valuable tool. But it should complement, not replace. To structure the right balance, we need to use AI as a learning tool. Developers should always review the AI generated code to understand the why and the how behind it. So deliberate practice is a key. Maybe a code might work AI generat code, but you should, as a developer or engineer, you need to understand the why and how behind, that specific code. So deliberate practice a key solving problems without ai critically. To maintain the strong problem solving is case. Do not just jump on the copilot chat, g PT or rec cascade from the venture. Just try to understand the error, on the terminal. I. Critical, review of AI generic code is essential maybe with a team member or by yourself throughout development. So treat AI like a junior developer that is there for you forever, right? So make sure that you actually go and verify the work of the junior. Now, finally. Please use AI as a upskilling tool. Learn new techniques and programming patterns from AI suggestions, but do not blindly accept them because there is an idea of HA hallucination in ai. So make sure you actually verify and understand the whatever AI is talking about. I personally believe that. These new innovative tools that we are using on a daily basis, they actually affect our feeling and impression toward our day-to-day job. So on positive side, AI removes the tedious fork and tasks and prevent burnout. That's great. But on the negative side, too much reliance on AI can, it would lead to disengagement in a way or boredom or. Job insecurity. So it is important to have that nice balance and, continuously work on yourself as an individual and learn the creative thinking and let the AI do the tedious stuff. So I. To keep ourself engaged. I believe we need to keep the work, challenging and meaningful. And, other day I told my colleague if we need to find a meaning in our work. So we cannot just blindly create a, I'm just giving an example, API endpoint or a new button in front end or new AI model version. No, we need to find a meaning and the why behind of every decision. Now encouraging ownership is another point where AI outputs are drafts that developers need to refine rather than accept them straight away. So it would keep the developers in control. And lastly, a growth mindset helps developers see AI as a tool for learning, not as a thread. Think of AI as a, your teacher where it's available, right? You write and it text backs, instantly, but you wouldn't want your supervisor or teacher to do the job for you. So you need to understand the concepts and, go deeper the why and hows of the, technology. Rather than just use it as a pilot to ensure developers stay competitive in AI driven world, companies must actively invest in upskilling and reskilling programs. Major tech companies have already taken steps to address this. Amazon, for example, launched a 1.2 billion upskilling 2025 initiative aimed at retaining a workforce in software engineering, ai, and cloud computing. At and t has committed 1 billion, to its feature ready retaining program, which provides, training in ai, data science and software engineering and software development. Many companies, including IBM, Google, Microsoft, have also established AI development, ensuring workforce remains skilled and adaptable as technology evolves. In this non-technical session, we have explored the negative impacts of ai. It is very important. you are not a software engineer. Even if you are not a technical individual. It is very important to balance out your AI usage. Use AI for efficiency purposes. I. But continue practicing your core skills. It can be coding. It can be accounting. no literature writing a poem. I don't know. Make sure that you do practice your key skills. Number two is it is very important to learn continuously. Always encourage yourself for new projects, for new ideas and challenges. Another point would be psychological engagement. Now. Keep work meaningful and challenging. just for the sake of it. Find meaning and challenge itself, and then start working on why and then how. And lastly, for business owners, they should invest in, upskilling and reskilling programs. It is very important to make the non-technical and technical staff aware of AI and the impacts that it is bringing. So this is the, understanding that I have in terms of the impact of ai. And if you have any questions, just let me know. Thank you very much. have a great and wonderful day.
...

Huseyn Gorbani

AI Developer @ Rider Levett Bucknall

Huseyn Gorbani's LinkedIn account



Join the community!

Learn for free, join the best tech learning community for a price of a pumpkin latte.

Annual
Monthly
Newsletter
$ 0 /mo

Event notifications, weekly newsletter

Delayed access to all content

Immediate access to Keynotes & Panels

Community
$ 8.34 /mo

Immediate access to all content

Courses, quizes & certificates

Community chats

Join the community (7 day free trial)