Jack Matusiewicz

Want to build your first audio plugin but don't know where to start? We've partnered with pointblank Music School and Plugin Boutique Ltd. to bring you "Basic Audio Programming" - a 4-hour hands-on course developed by Dr Chris Nash. No coding experience required...and by the end, you'll have built a JP8000-inspired SuperSaw synth capable of classic trance sounds. What you'll create along the way: 🎚️ A gain effect (fundamentals) ⚡ A clipping distortion (conditional logic) 🎹 A subtractive synthesizer (code objects) 🪚 A SuperSaw synth (arrays and loops) What makes it different: 60+ visual animations explain audio and computing concepts before you touch code. And it runs inside Klang Studio - a C++ IDE that lets you code plugins directly in your DAW. If you've been curious about what's happening inside your favorite plugins, this is your on-ramp. Find out more: https://lnkd.in/edaQeiK6

173

15 Comments

Unpopular Opinion: Experienced developers who '10x' themselves using AI are full of it. Sorry, I literally do not believe you. Of all the posts I've read of experienced SW developers who claim AI has made them '10x' more productive (and yes there are LOTs who literally say this) - not one has shown any evidence . The best I have seen is people who say "I finished this project in X days with AI and it WOULD HAVE taken me 10X days without AI'. Oh ? and how do you know that? One of the most important thing I have learned from 40+ years of SW development - 'we' suck at estimating how long things will take. Really Really Badly. Finally people have started doing scientific-ish studies to validate these claims. First result hot-off-the presses - 5+ year experienced SW developers given random real tasks performed 19% slower using AI then not. AND - most interesting, they believed they did it 20% faster. https://lnkd.in/gYSW4C8i

1,245

508 Comments

Looks like it could be happening: "With Proton, the Steam Deck could already do the Windows-to-Linux part; now, Fex is bridging x86 and Arm, too. This stack is what powers the Steam Frame’s own ability to play Windows games, of course, and it was widely reported that Valve is using the open-source Fex emulator to make it happen. What wasn’t widely reported: Valve is behind Fex itself." https://lnkd.in/e_Wpd4U3

1

When "Being Right" Isn't Enough. Marcus, a capable mid‑level engineer, asked his engineering manager, Sofia, for an urgent 1:1. "I can't work like this. I now fear every commit I make because I know it will start another fight." Marcus had joined the team relatively recently and was still building confidence in cross-team conversations. He had been pulled once again into a never‑ending PR battle with Jules, a long‑tenured senior engineer from another team whose opinion carried a lot of weight. Jules was technically sharp and often made valid points on the code, but at times his reviews crossed a line and had become a source of stress rather than growth. This wasn't the first time. Other developers had experienced similar behaviour from Jules in the past. Sofia recognised the impact this was having. She knew psychological safety mattered just as much as technical quality. So she acted. First, she made sure Marcus felt heard and supported. She extended the 1:1, moved her calendar, and went through the PR threads with him line by line. Together, they labelled what was technically valid and what crossed the line. "Your reviewer is often right about the code, but sometimes wrong about how he's saying it, and both of those things matter." She ended the meeting by reassuring Marcus: "We will fix this." Next, she involved Jules’ manager. Together, they reviewed the situation and agreed on some simple principles: • Design preferences, patterns, and styles are negotiable. • Respect and psychological safety are not. They used this as a chance to raise the bar for collaboration across teams by agreeing new review norms: • No sarcasm. Focus on behaviour, impact, and the code. • Limit async back‑and‑forth: once a thread gets more than a few comments, move to a short call. • Establish code owners across teams as accountable decision‑makers. • Use PR review rotations so cross‑team reviews become structured collaboration. They shared resources on constructive review practices: how to give more effective feedback and which behaviours to avoid. Sofia also paired Marcus with another senior engineer as a mentor to help him strengthen his architectural thinking and communicate his design decisions more clearly. Psychological safety is part of engineering excellence, and it starts with leaders who stand beside their engineers. #PsychologicalSafety #PeopleFirst #EngineeringManagement #CodeReview #HealthyTeams

1

building ai features that work in demos is easy building ones that work consistently in production is harder next thursday, my superstar colleague Tynan Byrne reveals how we ship enterprise grade ai features to customers like spotify, monzo at Omnea come through to learn how to iterate methodically and know whether changes are genuine improvements Thursday 5th February - 6pm Wave Talent HQ - Old Street Sign up here 👉 https://lnkd.in/eGXhs2Us

12

The 'Unreal' barrier is one I've seen a lot. If someone is an experienced engineer, especially talking in the double-figures of experience, then Unreal is 'just another engine' or 'just another editor'. Don't pass over hires (especially if they've got years custom engine experience) with blanket requirements around specific toolsets.

49

17 Comments

I've seen on a few posts lately on LinkedIn and I have to wonder.... What's the big deal with using AI to generate your CV etc? A CV is basically just a career fact sheet imo, so as long as the information is correct and accurate who cares? Not to mention everyone (Hiring managers, recruiting agencies etc) are using AI themselves to process CVs anyway? 🤔

19

1 Comment

We've been told of a coming apocalypse for software developers in the past - the advent of compilers meant that we didn't need programmers any more, 4GL, Visual Basic and a whole heap of no-code and low-code tools that seem to be on the scrap heap. In the meantime, C++ and C occupy the #2 and #3 spots in the TIOBE Index. I've been reimagining the future of development in the face of AI. Our day-to-day practices in software are very human-centric, and are going to change a lot, but most radically I predict in the job of writing software. Just as compilers meant that you didn't write 'proper software' you just used a tool like COBOL that abstracted it all away from you, agentic AI like opencode.ai and Claude is going to provide another layer of abstraction. My work today is in refining tickets, working on tests with the AI and the automation of the software feedback cycle, common engineering manager tasks. I don't write much code, but I still read some of it, but the job is to prevent the AI from making too many mistakes. The AI is capable of writing good code, and it is capable of making all of the mistakes you've probably seen for yourself, but saying that AI isn't ready is denying reality. If the AI is trashing your project its because you've let it, you are still trying to use it as a tool. You have to lead it, and that means that you have to rethink what it is you need to have to get results. As an industry we've done this lots of times for human teams, inventing sprints, tickets, kanban boards. How software requirements are defined has changed radically over my 30 years in software. Equally, how we manage teams has transformed from a group of individuals working in isolation to sprint teams. We've shifted-left, adopted DevOps, split platform engineering from product engineering and a myriad of other practices that are commonly accepted now.  The next stage is redefining all of these practices to improve how AI is writing software instead of the human. Tickets will need to be narrower in scope and more refined, but an AI will help to do that as well. Code review processes will need to be more automated, with AI code review and SAST as for every ticket. Processes we have that reduce the burden on human developers (linting, I'm looking at you) will become less important, and processes that reduce the burden on AI will be invented and adopted (I've had a few ideas). I'm rebuilding the SDLC, redefining what a sprint team looks like, going back to first principles and reimagining ways to solve problems. This will be big business soon, there will be a billion dollar company in whoever writes the best tools in the space where I (and I do mean Claude) am hacking together patches in Python. Developers, I urge you to rethink your roles. If we don't adapt to this new mindset we will not survive in an AI-driven development world.

17

2 Comments

Thoughts on Go after using it recently: - not being object orientated is a big plus. One of my least favourite parts of C#/Java - feels right at home coming from a typescript background - concurrency built into the language feels very nice - wish the type system had some more advanced features like discriminated unions - auto garbage collection makes the language very easy to use  - the fact that it compiles to a binary so no runtime is needed is a great advantage - seems like it's a language in between rust and C# which makes it a great tool for a lot of projects - love that go has errors as values and you can see which functions can return an error ahead of time, maybe one day we'll get something like this into typescript ( kinda is with effect.ts ) - interfaces being implicit instead of explicit is interesting and I'm not sure how I feel about it yet

6

I’m not normally so active on here, but am just really saddened about the SBG/Flutter redundancies. I agree with so much of the sentiment already posted by friends and former colleagues - SBG was THE place to be in Leeds and the wider North region for such a long time. Obviously change is inevitable but it doesn’t make it any less sad and I feel for everyone currently affected. I thought I would follow-up on my earlier post and offer some tips to anyone now looking for work for the first time in potentially a long time - albeit still obviously happy to chat if anyone wants to reach out directly. 1. Don’t panic. You’ll be ok. I would hope Flutter will look after you which will give you some time. 2. For now, and throughout your search, look after yourself. Take time off, stretch your legs. Lean on friends and family if you can, offload! 3. Persevere and stay positive. The right thing for you is definitely out there! 4. Use AI, but use it well. - Ask it to review and critique your CV (albeit redact personal info before pasting into ChatGPT etc). - Ask it to compare your CV to a particular job description. - Ask it to act as the hiring manager for a particular role and perform a mock interview. - DO NOT use its output verbatim. Write everything in your own voice, that’s the human part. AI-generated text is so obvious. And no - this post is not that! 5. A good CV is definitely important, so make sure it’s up-to-date. Don’t just list generic skills, rather what impact you made and what value you can bring. 6. Update your LinkedIn profile and make sure it’s consistent with your CV. 7. Upload your CV to jobsite.co.uk - recruiters definitely do use this in my experience. 8. Know what you’re good at (and what you’re not); it makes recruiter calls easier. 9. Don’t rely on making applications through LinkedIn. Complete waste of time IMO, you’re much better off going direct through company career websites, recruiters, hiring managers if you can. 10. Be proactive. Use your network if you have one. LinkedIn, former colleagues, tech meet-ups, etc. And/or reach out to recruiters directly. Get lots of calls scheduled, sell yourself, be enthusiastic. 11. Be hyper-organised. Use a spreadsheet with something like task, status, date, and notes columns. Add a new row for any recruiter contact or role applied for, and update the notes after any communication and set the date for whenever a follow-up is scheduled, or when you plan to chase. 12. When you land interviews, prepare. Do all the usual stuff - research the company, practice situational interview questions (and STAR method answers), prepare your own questions for the interviewers, etc. Again, AI tools can be quite good at helping with this. Hope any of that is helpful and I’m more than happy to chat as well. DM if you want any CV sanity checking or anything like that.

100

4 Comments

🚀 Last week, Will and I gave a live demo at the AI Tinkerers London meetup titled: “Turn Any Agentic IDE into Lovable/Bolt with Just One MCP.” We showed how simply adding an MCP server to Cursor—a coding agent—transformed it into a fully autonomous DevOps agent (or maybe... VibeOps agent? 🤣). It could create GitHub repos, deploy to Vercel, and configure DNS records on Cloudflare—all from a single prompt, hands-free. But what surprised me most wasn’t the success of the demo—it was what it revealed about my own outdated assumptions. 🤔 That moment made me rethink how I stay up to date with LLMs and agents. I realized my mental model was lagging behind the true circle of competence of the AI agents. So I wrote about it in a blog post: ✨ “When Your Mental Model Is Two Versions Behind the LLM” (link in comments 👇) Here’s the TL;DR: 🔁 Treat your mental model like a CI/CD pipeline Use personal evals to keep your “mental OS” up-to-date with each new release. 🚫 Avoid premature constraints on agents Test broadly first—don’t over-engineer for limitations that may no longer exist. ⭐ Evals are your North Star Map the agent’s true “circle of competence” before applying layers of control. 🌱 Grow with the models Ship lean V1s and let future LLMs lift capability. Build for the curve, not just the moment. 🔧 Shift to an AI-native paradigm AI is probabilistic and adaptive—our design thinking should be too. If you’re building with LLMs, I hope the lessons I learned can be helpful to you as well. 📽️ (Bonus: there’s a short demo clip in the blog as well.)

28

9 Comments