Duncan Anderson’s Post

How to Stay Relevant as a Software Developer We’re Drowning in AI… Now What? For years and years, we joked about robots taking our jobs one day. “It’s coming,” some warned. Fast-forward to a few years ago when AI exploded seemingly overnight. “It’s here,” those same people said. Enter budget cuts, mass lay-offs, and software developers the world over echoing a similar sentiment: “What now?” Here’s the good news: Brands are finding that swapping humans for AI might’ve been premature. Here’s the tricky part: Software developers still must choose to evolve or drown. How to Make Yourself Invaluable as a Software Developer Some of us have learned, perhaps the hard way, that while AI is incredible and will only get better, there are some things that will simply turn out better if a human being manages them. However, we’d also be foolish to deny that things have changed tremendously. I work with developers every single day, and I wanted to better understand what this evolution has meant for them. What can software engineers do to make themselves irreplaceable in the age of AI? I did my homework and also spoke with Luce Carter, Developer Advocate at MongoDB and Microsoft MVP, who first told me, “AI is not going away, especially with tools like Copilot. One of the best things you can do is learn a skill called prompt engineering, knowing the best and most effective way to ask the AI for what you want that will produce the most useful results.” Here’s what else I’ve learned. 1. Solve a Problem We’ve seen that AI is sometimes great for taking over monotonous, manual tasks — for instance, sifting through or generating code. Image by Luca Bravo But you, the software developer, still have an upper hand: You see the bigger picture. The end goal in mind. The target you’re aiming for. This is a skill unique to you. There are countless examples of how this can work. One way I’ve seen my teammates elevate the application development process using AI is via vector search, which understands the meaning and context of unstructured data, which it then transforms into numbers. How does this solve a problem? Well, it allows them (and/or their users) to more efficiently query data. This is a great example of humans and AI working together in harmony for the greater good. Remember, AI technology might sound cool but is a total moot point if we don’t use it to solve a problem and make people’s lives easier. If you’re not sure if your work is solving a problem, take a step back, look at what you’re doing, and ask yourself, “So what?” If you can’t come up with a clear answer, there’s more work to be done. 2. Make Way for Strategy Alright, you’re using AI, in some capacity, to solve a problem. You’ve found ways to leverage the technology to save you time. Now, you’re going to use that newly freed up time for… what, exactly? Scrolling on TikTok! Amazon shopping! Getting a snack! No, no, and yes because snacks are life, but then get back...

AI is not good software. It is pretty good people. Paradoxical ? We want our software to yield the same outcomes every time. So, we ensure that software systems are reasonably reliable and predictable. Large Language Models are neither of those things, and will absolutely do different things every time. They have a tendency to forget their own abilities, to solve the same problem in different ways, and to hallucinate incorrect answers. There are ways of making results more predictable, by turning down the level of randomness and picking a known “seed” to start, but then you get answers so boring that they are almost useless. Reliability and repeatability will improve, but it is currently very low, which can result in some interesting interactions. We also want to know what our software does, and how it does it, and why it does it. We don’t know any of these things about LLMs. They are also literally inexplicable. When you ask it why it did something, it is making up an answer, not truly reflecting on its own “thoughts.” There is no good way of understanding their decision-making, though, again, researchers are working on it. Finally, we should know how to operate a piece of software. Software projects are often highly documented, and come with training programs and tutorials to explain how people should use it. But there is no operating manual for LLMs, you can’t go to the world’s top consultancies and ask them how to best use LLMs in your organization - no on has any rulebook, we are all learning by experimenting. So the software analogy is a bad one. You should actually treat AI as people since that is, pragmatically, the most effective way to use the AIs available to us today. What tasks are AI best at? Intensely human ones. They do a good job with writing, with analysis, with coding, and with chatting. They make impressive marketers and consultants. They can improve productivity on writing tasks by over 30% and programming tasks by over 50%, by acting as partners to which we outsource the worst work. But they are bad a typical machine tasks like repeating a process consistently and doing math without a calculator . So give it “human” work and it may be able to succeed, give it machine work and you will be frustrated. And, of course, the AI still lies, makes mistakes, and “hallucinates” answers. But, again, so do humans. I would never expect to send out an intern’s work without checking it over, or at least without having worked with the other person enough to understand that their work did not need checking. In the same way, an AI may not be error free, but can save you lots of work by providing a first pass at an annoying task. We need to decide what tasks we are willing to delegate with oversight, what we want to automate completely, and what tasks we should preserve for humans alone. #Genai #LLMs https://lnkd.in/gzEtUqvr

Performing mundane tasks and reading through uncommented, messy code can be quite frustrating for developers 😖. With the advent of AI, why not utilise it to enhance efficiency? Explore the pros and cons of incorporating AI in development; click the link to learn how to protect your data🛡️ while using AI tools🚀.

The results are in. 📣 See how developers are using & viewing AI in the 2024 Docker AI Trends Report: 🧑💻 64% of developers report using AI for work 👍 65% view it as a positive option 🤖 23% of respondents see AI as a threat to their jobs Learn more:

"AI is not a trend" but a fundamental shift in software development. Docker's AI Trends Report 2024 shows key findings on how AI is impacting and improving development practices and how there is an increasing reliance on AI in everyday practices as the perception of it changes positively. Docker is fueling the mission to support developers with powerful solutions and provide the necessary tools to increase productivity and a richer experience. #ai #docker #development #developers

According to Stack Overflow’s 2024 Developer Survey, developers are embracing generative AI as a powerful ally rather than a threat to their jobs. With AI tool usage rising from 70% to 76%, and 81% of developers noting increased productivity, it's clear that AI is enhancing their work rather than replacing it. While trust in AI's accuracy is still a work in progress, and misinformation remains a key ethical concern, 70% of professional developers see AI as a supportive tool rather than a job-stealer. Overall, developers are optimistic about AI’s potential to improve their work and drive innovation. #GenerativeAI #SoftwareDevelopment #Innovation

🚨 Warning: The Dangers of Misused AI in Software Development 🚨 Note: I did use AI to rewrite this post as I think I needed to take all of my notes, the interaction I had with the "director of business development" and some research I had done PLUS a conversation I had with my very seasoned team. This is where you can use generative AI to your advantage AND not for the following reason: Today I had a deeply troubling interaction with a platform (PM me for the actual name of the company) that promises to simplify project scoping for non-technical users with AI-driven tools. While their pitch sounds appealing, the reality is far from what they advertise. Here are the key issues I encountered: 1️⃣ False Expectations and Misleading AI Company X uses AI to create user stories and project documentation, but in my experience, it bypasses the most critical aspect of software development: human expertise. AI-generated documents don't account for the nuances of a client's needs, which could easily lead to scope creep and misaligned expectations. Developers and agencies are left to clean up the mess, often without proper guidance from the platform. In fact, the biz dev couldn't even explain how scope creep works on the platform - btw at that funny agency we have figured out how to avoid it altogether. 2️⃣ Pay-to-Play Model for Developers Here's the kicker: developers must pay a monthly subscription just to access project leads. While it might sound like a typical lead generation platform, this model doesn't offer fair value exchange. In fact, it risks exploiting smaller agencies and developers who may be desperate for work. It's a classic case of "pay to play" that puts a strain on developers who are already fighting for limited opportunities. Horrifying to think you have to develop what they say or you don't get work even though the mistakes will all be on you. 3️⃣ Lack of Real Scoping and Advice When dealing with large-scale projects, clients need guidance and advice from experienced professionals, especially when it comes to complex tasks like architecture, user stories, and technical scope. Instead, company x's model just opens the floodgates, allowing clients to input whatever they think they need without any expertise to steer them in the right direction. This only sets developers up for frustration when they realize the client's original scope isn't realistic or feasible. So, what's the takeaway? While AI has a lot of potential to streamline processes, it should never replace human expertise—especially in something as nuanced as software development. If you're a developer or agency considering company x, beware of these false expectations, the pay-to-play model, and the lack of real support. Let's raise awareness and ensure that tools like these are held accountable for their practices. #AI #SoftwareDevelopment #TechEthics #Scoping #WebDevelopment #LeadGeneration #PayToPlay #ScopeCreep #FalseExpectations #DevCommunity

Let me guess. You’re introducing AI into your software development lifecycle, but instead of magic happening, it feels like you’ve dropped a wrench into a finely tuned engine? Yep, been there. The issue probably isn’t the AI, it’s how you’re trying to fit it into a team that has no idea what it’s supposed to do. You can’t just sprinkle AI over your SDLC and call it a day. You need to rethink how your team works together. If you think this is something you can “automate and forget,” well, you’re in for a shock. Here’s what needs to happen instead: Strategy first, automation second: Before you even begin plugging in AI tools, ask yourself, “What specific problem am I trying to solve?” If you don’t know, neither will your AI. Define measurable goals, don’t just “automate because AI is cool.” Data is your new lifeblood: Your AI model is only as good as the data it’s fed. You don’t need random streams of garbage data from a decade ago, but you do need data that is clean, relevant, and recent. Establishing strong data governance is critical. Treat your data team like royalty, they’re the ones feeding your AI “baby.” (PS this is where most organizations fail as the cost of cleaning up decades of problems isn't free). Get the right people in the room: AI integration isn’t a one-person (or one-department) job. You need a cross-functional team. Yes, that means roping in the data scientists, domain experts, software engineers, and even the product managers. These different brains together? That’s how you’ll actually get AI working for you, not against you. Explainability matters: Trust me, when AI starts making decisions you can’t explain, panic sets in fast. You don’t want your code base to feel like it’s being haunted by a rogue algorithm. Use explainability (XAI) tools like LIME or SHAP to break down how your AI is arriving at its conclusions. Continuous learning is non-negotiable: AI models need regular updates, tweaks, and continuous learning. Sorry, you can’t just set it and forget it. Unlike a fine wine, AI doesn’t get better with age unless you keep feeding it fresh data and reviewing its performance. So, next time you’re about to blame AI for your dev problems, ask yourself if you’ve built the right team, cleaned up your data, and actually understand what the AI is doing. If not, you’ve got bigger problems than a buggy line of code. What’s been your biggest challenge in getting your team to fully embrace AI integration, and how have you tackled the skepticism or confusion that comes with cross-functional collaboration? https://lnkd.in/ejnXqikr #AIinTech #SoftwareDevelopment #AIIntegration #AIAutomation #TechInnovation

New research from Stack Overflow reveals that AI tool usage among developers increased to 76% in 2024, up from 70% in 2023. 81% of developers cite increased productivity as the top benefit of AI tools and 70% don’t see AI as a threat to their jobs. In fact, individuals believe AI is democratising coding, meaning the number of developers is increasing rather than decreasing. Yes, yes, I think everyone understands AI = helpful. But what's the next step for the everyday developer? From what I hear, is using sandbox playground environments to run such codes. What do you think? https://bit.ly/3WmvjXY #keepupwiththenews #keepupwithyourskills #pluralsight #acloudguru