https://dankim.org
50 posts · 4 Votes
Tech 56% · Life 24% · Culture 18% · Writing 2%
Subscribe via RSS
This, so much. Something seems to be lost on my peers today: it’s still easy to not use AI. The food we eat, clothes we wear, and every electronic device we touch may embody innumerable injuries to the world, and all this is inescapable. Eschewing AI is one thing that we can actually do to live out ethics that affirm values of human and environmental rights. It’s almost a gift! Just use a computer the same way you did three years ago!
This has already been making the rounds but allow me to add my favorite part that I actually lol’d at… It turns out that chatbots are good at chat. It claimed to have implemented authentication. Actually, it had basically written: function authenticate(user: string): bool { return true; } I am fairly sure this is not compliant with the WebAuthN/Passkeys specification, though I haven’t read it to be sure.
God I hate LinkedIn with a passion. Just rage wiped all my activity there. There’s a semi-subtle undercurrent of toxicity there that’s easy to miss if you’re not looking for it (humble brags, hype trains, AI bullshit among them). They interleave some nice stuff, like your friends’ real accomplishments, to keep you hooked. But for the most part the content just eats away at your brain and you kinda don’t even realize it until you actively take a step back and take a hard look.
I wrote a document today. I used my brain and no LLM was consulted. It took me 10 minutes to write and it’s 3/4 of a page long, yet it has all the basic info we need to get started. I’ll be able to reason about it, reference it, and talk about it succinctly without issue BECAUSE I WROTE IT.
When asked how Meta is tackling these privacy controversies, a spokesperson for the company told Vogue Business: “We have teams dedicated to limiting and combating misuse, but as with any technology, the onus is ultimately on the individual to not actively exploit it.” What an absolute load of horseshit to push responsibility back onto consumers. They’re your creation, assholes!
Important stuff! tl;dr: you might be an excellent land athlete or even a great pool swimmer, but swimming in the ocean or a lake is very different and much harder than you think. 2026 so far: 30 deaths, 20,000 rescues. Be careful out there, friends.
I talk about a bunch of things here with my good pal Huyen (Kotlin Multiplatform, Android, the web) but arguably the most important thing we chatted about was the role of AI/LLMs, how it’s a tool and should be treated as such, and why the human side of engineering still matters the most. https://www.youtube.com/watch?v=vTNEs3EUSJo&t=682s
It’s true at work as much as it is at home – model the behavior you want to see more of, fight for what you believe in, and hope for the best. ✊
Easy way to sniff out a bad leader: they use “I” language for accomplishments and achievements and “we” language for accountability and unfavorable outcomes. Literally the opposite of what you’re supposed to do and such an obvious, blatant red flag.
People and the culture they build are your irreplaceable competitive advantages, period. If you’re so concerned about building a “moat”, that’s what you invest in.
When it comes to managing people (and by association, the success of your team), your most valuable currency is trust. You get a little “starter money” at the beginning, but from that point on it’s a full time job to earn more of it. It takes time and a concerted effort, but it’s one of the most important assets you can develop as a leader. If people trust you, you can get through some seriously hard times together.
I cannot overstate how important it’s been in my life to build friendships at work and in developer communities. I don’t mean just lightly getting to know colleagues, I mean real, actual friendships. More than anything else, those are the single most valuable thing I’ve carried forward from any job or conference. 💙
This post by Jonas Downey is a must read. He talks about all the feelings of being in tech right now with AI bearing down on us. But most importantly, he leaves us with a sense of hope and a reminder about our responsibilities. We, the humans, can absolutely force change and course correct this largely unprincipled direction. We just have to do the work (as always). Let’s remember that tech movements are not inevitable because some hype bros said so. Just look at the Metaverse or NFTs. It…
I’ve already had a long, fruitful career. I’d like to keep it going, but ultimately I’m not here to carve out more for myself. I’m here to help make it right for the folks coming up.
Sure I totally believe that white paper you wrote about AI adoption and related concerns, company that sells products and services that help with AI adoption and related concerns.
You should read this. Adam might be one of the few brave enough to say it, but everyone except hype bros is feeling it. These LLMs are prediction machines. They are text generators that are ultimately a bunch of fancy statistics, trained on the years and years of dedication by brave engineers willing to learn and build in the open. Building in the open meant we were not gatekeeping technology, but creating tangible examples for young engineers to explore, understand, and learn from.
A VC NFT/crypto bro said he “100% disagreed” with one of my posts that was critical of blindly following the AI hype. I for one am shocked.
This piece from Anil Dash remains my absolute favorite take on AI. It’s fair and pragmatic, and in a word, NORMAL. I think I read it every day and twice on days when I feel like I’m losing my mind amongst the AI hype. Technologies like LLMs have utility, but the absurd way they’ve been over-hyped, the fact they’re being forced on everyone, and the insistence on ignoring the many valid critiques about them make it very difficult to focus on legitimate uses where they might add value.
I think a lot of times people mistake information for culture. Sending a bot off to summarize Slack for you can get you a lot of data, but it certainly doesn’t get you any closer to understanding your people. And in any successful organization, the people are everything. No LLM is going to be able to read tone with your judgement and context, no matter how many instructions you feed it. Period. End of story.
A little while back I tried fully embracing the “I want it my way and I want it now!” AI mindset just to see what would happen. I wanted to see first hand if any of the hype held (spoiler: it didn’t), so I slapped together a handful of apps with LLMs to fulfill what I considered to be, at the time, niche needs that other apps didn’t do perfectly for me.
People are going to start running with this idea that 1:1s aren’t needed now because AI Jensen says so. Folks who are already weak leaders will buy into this as an excuse to cancel all their 1:1s, but that would expose a fundamental misunderstanding about what they’re for. 1:1s are NOT for you, they’re for your direct reports. It’s a chance for them to get guidance and help. To get ideas. To develop their careers. To vent. To complain. To feel supported and understood. To develop trust. To…
Vibe coding is the equivalent of me sketching a drawing of my dream car, it materializing in front of me, and then proclaiming “I MADE A CAR IT’S 80% DONE”. And then the first time I drove it the wheels would fall off, the engine would explode, and I would drive off a cliff. Seems fine.
My LinkedIn timeline today: The Engineering Manager role is dead Wake up call for engineers — clean code doesn’t matter, only results Gonna be fun to come back to those posts in a couple years. 🚮
I can’t believe I have to say this, but just because some dude says they shipped something in a week using AI doesn’t make it true (or good).
Building a full stack webapp entirely using Jetbrains frameworks and Kotlin has long been a dream of mine. If you’re like me (a longtime Android and backend developer), you want to write everything in Kotlin including the front end, but doing that has historically been the hardest, trickiest part. While there are some frameworks that can help bridge the gap, it’s always felt a little disjointed when integrating them with the rest of the stack.
Pretty much every company has AI bros, we know that. And if we’re lucky, there are also some pragmatic, vocal folks fighting for balance and humans. But there’s also a group of folks caught in the middle who are just trying to get by. They’re leaning into AI because, quite frankly, they’ve been told they have to. They literally don’t have a choice. For any number of good reasons, they need to keep their jobs, so they’re not in any position to publicly push back on their AI overlords right now.…
I’ve heard a version of the phrase “This isn’t what I signed up for” a few times this week, from vastly different folks in tech. People in very different companies of different scales and in different disciplines. But all were expressing the same basic sentiment, which was more or less “Working with AI all day and not other humans is not what I wanted to be doing, but here I am.”
Well made software products and well run product companies will make it through the AI hype cycle, no problem. Pull up any vibe coded project that’s “70% done” or “something that took me 10 minutes to build”. Then have that person do the calculus on everything it would take to close that last 30% to productionize it, to find a market fit, to acquire customers, and to actually make money off it. Oh yeah, don’t forget to remind them that they also get to maintain it forever. Sit back and watch…
Starting a post with anything remotely like “I’m in the top 1% of AI users…” is not the flex you think it is. First of all, there’s no way you could possibly know or quantify that. Secondly, get a life.
The three phases of execs with AI brain rot: Wow, look at how powerful these hammers are. I have so many hammers! Why are these craftspeople only using one hammer? Surely they can use 12 hammers at once and also the hammers can self organize anyway. We don’t need all these people, let’s lay them off and save money! I don’t understand why all these hammers aren’t working.
