Marko Mihovilić

I am very impressed by the “Machines of Loving Grace” essay by Dario Amodei. It describes possible changes in 5-10 years after a powerful AI is developed and massively deployed. https://lnkd.in/dERCG3Mk I like the title. It’s from the “All Watched Over By Machines Of Loving Grace” by Richard Brautigan. “I like to think (it has to be!) of a cybernetic ecology where we are free of our labors and joined back to nature, returned to our mammal brothers and sisters, and all watched over by machines of loving grace.” I also like that Dario doesn’t say how exactly we can build the powerful AI, but leaves the pointers to scaling laws and the bitter lesson in plain sight. There are basically two kinds of people in generative AI. Those that internalised the scaling laws / the bitter lesson, and those that haven’t. The Bitter Lesson by Rich Sutton is written in March 2019. If you didn’t read it yet, go and read it - https://lnkd.in/drZzfu99 It helps to understand why companies build enormous AI datacenters powered by compact nuclear reactors.

8

2 Comments

There is a new open weights model on the Hugging Face, called MiniCPM V2.6 by OpenBMB. Of course, I tried it locally on my personal MacBook Pro and the HuggingFace space. I used a receipt from two days ago as the input image and asked some questions about the image, including an attempt to extract all the text from the receipt image and format the output in a JSON format. The model wrote the JSON format correctly, but it was not in the video. Perhaps it was because I didn't provide a few-shot examples for the model. Considering the model's size, it's an excellent model to tinker with. I'm not sure yet about running it on a production scale. Primarily because I didn't have enough time to explore the limits of this model. I'm interested in running it natively on the MacBook using the MLX framework instead of using PyTorch with the transformers module. Anyway, it was my first experience with the model. It's a good first impression. Links to the video: - Running locally on my MacBook Pro: https://lnkd.in/g7Du4jyw - Trying on the HuggingFace space demo: https://lnkd.in/gFXs2Rf2 Try it online at https://lnkd.in/ggDHQp7D See the model card at https://lnkd.in/gZ5PR4i4

94

1 Comment

Very cool and simple course how to get AI agents running within minutes.

3

I didn't expect to say it this soon, but I'm enjoying working with TailwindCSS. I have used it for the AutoProject webapp. And with the help of LLMs, writing TailwindCSS has become quite easy. I often just describe the design I want and ask GPT to come up with the TailwindCSS for it. It works well, most of the times. Just a month or so ago, I talked about how I disliked Tailwind for its very long and verbose CSS classes. Every html element in your code will have a long list of class names to get the required design. I still dislike it, but after trying it more, I felt the productivity gain is worth it. Prior to this, I always preferred writing my own CSS because I wanted the design to be very unique and customized to me. In hindsight, that's still good but it reduced my shipping speed and added unnecessary "figuring it out" and "testing" overhead. I am continuing to write my personal CSS + Component library to factor in that uniqueness, but in the meantime, tailwind looks like an excellent tool to iterate very fast and build great looking UIs. With a stable Tailwind version 4 releasing soon, it looked like the right time to pick it up. In version 4, we can expect: - 10x faster builds - 35% smaller in size - Use of lightning CSS (I'm using this for my personal CSS library too!) - A custom parser that's 2X faster than PostCSS. (They literally designed new data structures to achieve this!) If you have been a skeptic like me, to Tailwind or CSS libraries in general, don't sleep on it. I'm thinking of migrating even TechFlix to TailwindCSS.

26

3 Comments

Not only is it a super sexy smol MoE model, but all of its *data* is fully available as well! Now *that* is what I call open-source. Probably needs a new name, to distinguish it from the models which are open-source code (everybody), and completely freely useable (i.e. totally permissive license: much fewer players), *and* now also all the training data (almost nobody).

19

🚀 Ever felt overwhelmed by the sheer number of tech tools out there? 😅 You’re not alone! I have been getting a lot of questions about how I approach learning new technologies, and I realized it’s time to share my insights! ✍️ In my latest article, I outlined my personal strategies for mastering new tech tools efficiently. 💡 https://lnkd.in/dxDTwA-C Whether you are a seasoned developer or just starting out, these tips can help streamline your learning process and boost your productivity. 📈Curious about my journey? 🌟 Check it out and let’s discuss how you tackle learning in this fast-paced tech world! 🤔💬 #Learning #TechTools #DeveloperJourney 

32

2 Comments

🎉 Exciting moment: One of our talented engineers just published a blog post diving into the Visitor Design Pattern, part of our ongoing Design Patterns Book Club series! 📘✨ For those new to it, the Visitor Pattern is a design pattern that lets you add further operations to objects without having to modify their structures. It’s especially useful in scenarios where you want to perform different actions across a set of related objects. Inspired by a presentation he shared with the team, this content was created for both senior developers and those just getting started. It breaks down the Visitor Pattern in a way that feels approachable, practical, and impactful. Spoiler alert: code examples in Ruby! 😍 If you’re looking to deepen your understanding of the Visitor Pattern, or if you’re mentoring junior devs who could use a clear and detailed guide, check out the post on our blog: https://lnkd.in/eSwu39-R Kudos to Leonardo Vailatti Eichstaedt for doing an amazing job of breaking down complex concepts and for being a key part of our collaborative learning journey. More cool stuff coming soon, one pattern at a time!

41

3 Comments

Internet rumors... 😱 Tomorow will (maybe) be a big day! Tomorrow might also mark the return of 'work from home'... Can you imagine being in an office where everyone is speaking to their own assistant? 🤣😂 #VoiceAssistants #FutureOfWork #WorkFromHome #TechTrends #OfficeLife #SmartOffice #AIInnovation #DigitalWorkplace #RemoteWork #WorkplaceTechnology

4

Gave a talk at work on "Dependencies" today, trying to externalise and make explicit some of the background awareness that shapes my everyday practice as a software developer. It takes time for this awareness to develop. I remember the first time someone talked to me about how Gradle uses a topological sort over a graph of dependencies to determine a build order for tasks, and what a revelation that was. I remember the first time I was shown how to get IntelliJ to scan a Java project for cyclic dependencies between packages, and coming to understand why one might want to avoid them and how to unpick them if they arose. Now DAGs (directed acyclic graphs) are a comfortable piece of my mental furniture, but I don't know that more junior developers are often directly introduced to them - they might come up as a data structure with associated algorithms in CS courses, but is it explained that they're fundamental to the way compilers and build tooling assemble working software from many related pieces? Many developers are aware at some level that transitive dependencies can balloon out into bloat (and an ever-expanding attack surface for vulnerabilities), but how many can articulate this clearly? A large part of my practice as a Principal Engineer is mentoring individuals and looking out for knowledge gaps in teams. It can be hard to recover a "beginner's mind" and remember what it was like not to know (or to be only vaguely aware of) something that became second nature to you over a decade ago, but it forces you to make your implicit knowledge explicit - which also gives you opportunities to reflect on it. Teaching is learning: you go back and retread old paths, and are confronted again with long-forgotten questions. Like everyone at the moment, I wonder what the impact of LLMs is going to be on the way developers acquire the background awareness and instincts that guide accomplished practitioners. Will they help to make implicit knowledge explicit, or will they paper over gaps in knowledge and obstruct the process of learning?

12

2 Comments

"If you want something, ask for it." A simple yet powerful advice that I picked up early in my career and that has stuck with me ever since. Read all about it here: https://lnkd.in/dts4zq9k

6

Patience is not always a virtue. Well, at least, it isn’t always great for your career - especially when it comes to promotions/growth. If your manager is asking you to be patient about your promotion, push back and ask for specifics: * what exactly are the behaviors I need to demonstrate to be eligible for a promotion, and what are the upcoming projects that will give me the opportunity to do so? * what is the expected timeline for when I can be considered for promotion? If you’re not getting clear answers to these questions, really consider your options because it is unlikely your leadership is invested enough in your growth. Asking you to be “patient” is just their way of keeping you quiet for the moment. ADVOCATE FOR YOURSELF.

101

3 Comments

Comprehensive report on the state of the web, backed by real data. 21 chapters, some still WIP but definitely worth reading ➝ 2024 Web Almanac https://lnkd.in/ddkvqT3P

2

After more than 2 years of work (April 6, 2022!), my macOS app, PR Focus, is finally ready for public beta! If you are a #TechnicalWriter who works with engineers who occasionally forget to mention when something needs a documentation update, or... If you are a #Developer or #DocsAsCode technical writer who makes or reviews PRs, or is assigned work in GitHub... You might find PR Focus a useful tool to help you keep track of PRs across multiple repositories - all in a single dashboard. I've just finished adding more info and some App Preview videos to the website: https://prfocus.app/ I have been using it daily at work for over a year in its current incarnation, have had a group of colleagues using it to provide feedback, and am finally satisfied enough with the level of fit and finish to share it with others. I have gotten *so much* value from it, it has kept me learning and working through the queue of bugs and grown to the point that it might be quite useful to others. If you're interested in joining the public beta, you can join the TestFlight beta group here: https://lnkd.in/esRnx3ZJ

43

3 Comments

I ran across this post in the Changelog newsletter: https://lnkd.in/eCdjVekq - I like the name slop for unwanted AI-generated content. I love LLMs. They are part of my daily workflow, and I've been exploring them more seriously while on my sabbatical. I wouldn't misrepresent generated content without critical review as my own, but the web seems to be filled with more generated garbage (or 'slop')—content that looks fine at first but isn't accurate or helpful. LLMs (and other generative AI) should help us to make things of higher quality, not just churn out slop more quickly.

11

1 Comment

Really curious to see where this goes 🍿 My personal view is that #AI is not a “trend” and cannot be avoided. Companies and professionals should harnessed it and leverage it. It will (is it already? That’s another post…😬) enhance human productivity BY FAR MORE than any other tech disruptions of the last 40 years: more than the PC, more than the internet, more than mobile phones. We are in front of a piece of technology that can give #people the #FREEDOM of focusing on the most important thing™️ taking away the chores from the table. #embracechange https://lnkd.in/edQjvMmQ (Original interview on the NYtimes here behind a paywall https://lnkd.in/e3P6wFZF )

6

3 Comments

My thoughts about Google IO, innovation, the AI hype and "putting all your eggs in one basket". TLDR: 🤔 I think that AI is overhyped! 🔥 Video and image generation is wild though (not mentioned in the video) 🔎 Google IO is not cool anymore. What do you think? 🥚 DONT PUT ALL YOUR EGGS IN ONE BASKET! if they pull the plug on you ... you're done. #googleio #innovation #techstartupsuk #cto #techinvestor #fintech #ai

3

At Localyze we just finished our annual performance reviews cycle. This year I got feedback not only from my manager but also anonymous 360° feedback from peers and, for the first time, feedback from my direct reports. Of course, there are always areas and ways to improve, but overall I am very satisfied with the review. Nothing unexpected was brought up, and there was lots of positive feedback and a bit of actionable advice how to improve. Like many software engineers, I often experienced signs and symptoms of the impostor syndrome during my career. However, after more than 16 years in the industry, I finally feel confident about my technical skills and experience as a software engineer. But as a manager, I still have lots of doubts. For too long, I avoided doing people management because I thought it would be too hard and demanding for me having meaningful, helpful and sometimes difficult conversation with reports on a daily basis. And I have high expectations of what makes a good manager. Having integrity is an important aspect, I want to be recognized a transparent, trustworthy, and emphatic leader. Therefore, this short paragraph from my performance review was my highlight. It makes me very happy, and I am very proud that the people around me perceive me being empathic and see it as one of my greatest straights.  Because the feedback was anonymous, I don't know who to thank. But: Thank you very much, teammate. This paragraph means a lot to me, because it means that I am on the right track.

74

2 Comments