HNHacker News
TopNewBestAskShowJobs

xwowsersx

4,421 karma · joined June 6, 2012

email: xwowsersx [at] fastmail [dot] com
submissionscomments
xwowsersx··on Charlie Kirk killed at event in Utah
Really dislike this. It risks sounding like a justification, because even if it only means someone will inevitably react violently, its vagueness makes it read as though violence is excusable or natural.

Not so. People can and should endure rudeness, even disgusting behavior, without throwing so much as a punch.

Our entire legal system is built on the premise that violence is not the natural/inevitable outcome of incivility. Courts, contracts, and laws exist precisely to channel disputes, insults, etc. into nonviolent processes.

If violence were the automatic consequence of rudeness, there'd be no point in having civil courts/workplace dispute procedures/defamation law... or even law enforcement protocols in general. The system assumes that people can and must respond to incivility without physical aggression and it punishes those who don't.

xwowsersx··on Gleam Gathering 2026
Funny timing. I literally just started poking around with Gleam last night. Got it installed, I think I have the LSP set up, though I wasn't really seeing much autocomplete in nvim. Either way, I'm already pretty excited about it. I love functional languages, I like fast languages, and I like them statically typed. Gleam seems to check all those boxes, so I'm looking forward to digging in more.
xwowsersx··on Understanding Transformers Using a Minimal Example
So glad you shared this. Super accessible without diluting. Thank you!
xwowsersx··on We're Joining OpenAI
I'm speculating here, but my guess is that the urgency for OpenAI isn't that the Alex team has a "secret prompt." It's that they've already done the heavy lifting to make LLM coding assistants actually usable in a domain (Apple platforms) where OAI doesn't have integrations yet.

Acquiring them gives OAI:

- ready-made team that understands the IDE plumbing and the developer UX at a deep level

- head start in a platform ecosystem that's hard to crack

3. team that knows how to push the models into productized, developer-ready experiences

So it's not the prompts. It's engineering the scaffolding and UX around the model so it feels like magic to the user. That's what OpenAI is buying.

Also possible that Alex solved a lot of UX and integration challenges that could translate to other Apple contexts (productivity apps, design tools, even consumer-facing AI on macOS/iOS).

Without having used Alex myself (I don't do iOS or macOS development), I would guess that all the retrieval, context slicing, editor integration yada yada that they've built aren't necessarily unique to coding. Same scaffolding could support things like AI-driven writing, design, or general productivity in an Apple-native way.

xwowsersx··on Magic Lantern Is Back
Same :) I had in mind Groove Salad from soma.fm
xwowsersx··on Claude Sonnet will ship in Xcode
They might be referring to the plugin https://plugins.jetbrains.com/plugin/27310-claude-code-beta-
xwowsersx··on A visual introduction to big O notation
This is excellent. Very nicely done.
xwowsersx··on My experience creating software with LLM coding agents – Part 2 (Tips)
This lines up with my own experience of learning how to succeed with LLMs. What really makes them work isn't so different from what leads to success in any setting: being careful up front, measuring twice and cutting once.
xwowsersx··on AI is different
It's not a No True Scotsman. That fallacy redefines the group to dismiss counterexamples. The point here is different: when the thing itself keeps changing, evidence from older versions naturally goes stale. Criticisms of GPT-3.5 don't necessarily hold against GPT-4, just like reviews of Windows XP don't apply to Windows 11.
xwowsersx··on Do Things That Don't Scale (2013)
Something I heard recently on some podcast which really resonated was (paraphrasing): "Inertia is the biggest issue for a startup. The world doesn't like you, doesn't think it needs you... and you've got to invert that. You have to create momentum by scratch and the most important thing a founder does, by an order of magnitude, is invert inertia. Literally, in a physical sense. Like the world stays at rest and you've got to create momentum." With that in mind, it makes sense to do things that don't scale... because in the beginning, you're not trying to optimize a machine that's already running. You're trying to get the engine to turn over at all. PG's point is that scalable growth comes later; first, you have to manually crank the thing into motion. You do manual stuff not because it's efficient, but because it's the only way to get traction when no one's looking for you in the first place. And that's how you learn what actually works, how you build momentum one user at a time, and how you prove there's something worth scaling at all.
xwowsersx··on Convo-Lang: LLM Programming Language and Runtime
I haven't used DSPy this much, but as I understand it: this lang is more like an orchestration DSL for writing and running LLM conversations and tools, whereas DSPy is a framework that compiles and optimizes LLM programs into better-performing prompts...like DSPy has automatic improvement of pipelines using its compilers/optimizers. With DSPy you deal with modules and signatures.
xwowsersx··on Convo-Lang: LLM Programming Language and Runtime
Not sure I'm sold on this particular implementation, but here's my best steelman: working with LLMs through plain text prompts can be brittle...tiny wording changes can alter outputs, context handling is improvised, and tool integration often means writing one-off glue code. This is meant to be DSL to add structure: break workflows into discrete steps, define vars, manage state, explicitly control when and how the model acts, and so on.

It basically gives you a formal syntax for orchestrating multi-turn LLM interactions, integrating tool calls + managing context in a predictable, maintainable way...essentially trying being some structure to "prompt engineering" and make it a bit more like a proper, composable programming discipline/model.

Something like that.

xwowsersx··on I tried coding with AI, I became lazy and stupid
I think one of the reasons "coding with AI" conversations can feel so unproductive, or at least vague, to me, is that people aren't talking about the same thing. For some, it means "vibe coding" ... tossing quick prompts into something like Cursor, banging out snippets, and hoping it runs. For others, it's using AI like a rubber duck: explaining problems, asking clarifying questions, maybe pasting in a few snippets. And then there's the more involved mode, where you're having a sustained back-and-forth with multiple iterations and refinements. Without recognizing those distinctions, the debate tends to talk past itself.

For me, anything that feels like anything remotely resembling a "superpower" with AI starts with doing a lot of heavy lifting upfront. I spend significant time preparing the right context, feeding it to the model with care, and asking very targeted questions. I'll bounce ideas back and forth until we've landed on a clear approach. Then I'll tell the model exactly how I want the code structured, and use it to extend that pattern into new functionality. In that mode, I'm still the one initializing the design and owning the understanding...AI just accelerates the repetitive work.

In the end, I think the most productive mindset is to treat your prompt as the main artifact of value, the same way source code is the real asset and a compiled binary is just a byproduct. A prompt that works reliably requires a high degree of rigor and precision -- the kind of thinking we should be doing anyway, even without AI. Measure twice, cut once.

If you start lazy, yes...AI will only make you lazier. If you start with discipline and clarity, it can amplify you. Which I think are traits that you want to have when you're doing software development even if you're not using AI.

Just my experience and my 2c.

xwowsersx··on AI promised efficiency. Instead, it's making us work harder
Well, a few things to consider:

Investment capital isn't zero-sum...if/when AI generates outsized/large returns, it actually brings more money into the entire VC ecosystem. LPs who 10x their AI investments don't exactly hoard that cash, they reinvest and often diversify into other sectors.

Every major tech wave creates huge downstream opportunities. The internet "bubble" didn't just benefit search engines...it spawned e-com, SaaS, fintech, etc. AI is doing the same thing with robotics, new semiconductors, data infrastructure, and I'm sure other categories that don't even exist yet.

Also investors know they need portfolio diversification. Even AI-focused VCs are actively looking for contrarian bets in undervalued sectors precisely because there's less competition there right now.

Plus, AI advancement should (yes, I know there's hype and theses that may not play out) accelerate innovation everywhere else. Example: A biotech startup today has access to AI tools that were Google-only five years ago. This makes non-AI startups more attractive, not less.

We saw identical complaints during the dotcom era about "real businesses" getting ignored, but that period actualy coincided with growth across enterprise software, telco, and tons of other sectors.

xwowsersx··on UK backing down on Apple encryption backdoor after pressure from US
I mean, yes many policymakers still struggle with the nuances of modern tech, but claiming that "the needle hasn't moved" in 25 years is an exaggeration. In the late 90s/early 2000s, encryption debates featured lawmakers who barely understood email. Since then, there are committees focused specifically on tech policy, even some lawmakers with backgrounds in CS or cybersecurity... and far more nuanced public debates about encryption, surveillance, and privacy.

I recently listened to some clips from a hearing with questions about zero-knowledge proofs, algorithmic transparency, etc...this was pretty unthinkable two decades ago. Some agencies and legislative bodies also now have technical staffers and some advisory boards with technologists. So, yeah it it slow and sometimes frustrating, but it's not static.

xwowsersx··on What happens when clergy take psilocybin
Promising title, but the article felt hollow... all surface, no depth. Skimmed anecdotes without probing them, offered no real insight or new perspective, and left me with absolutely nothing I couldn't have guessed from the headline alone :(
xwowsersx··on Ask HN: How do I give back to people helped me when I was young and had nothing?
Your week?! This would brighten my entire month, perhaps even my year!
xwowsersx··on Show HN: Ikuyo a Travel Planning Web Application
My family and I have a trip to London/Paris in July with children. Tried Wanderlog and it's not great. Going to give this a try!
xwowsersx··on Changing Directions
So I totally get feeling disillusioned...tech has definitely caused harm, no question. But writing off the whole industry as just a force for bad is kinda missing the full picture. It's like only seeing the shadows and ignoring all the light.

Yeah, "surveillance capitalism" is real (though I might take issue with the terminology). So are predatory gig models and all the rest. But the same tech that powers that stuff also lets a kid in a village take a free coding class. It lets someone stuck in a war zone video call their family. It lets people in remote places talk to a doctor without traveling hours. It helps small business owners make a living without middlemen, gives a global voice to folks who used to be ignored, and helps disabled people live with more freedom.

Tech has also made a huge difference in:

- making communication and education basically free

- speeding up medical breakthroughs (including vaccines)

- Helping organize disaster relief and mutual aid

- connecting people to jobs, homes, and communities

- bringing corruption and injustice to light thanks to mass documentation

Of course there's a dark side. Every powerful system has one. But saying "this is all tech ever did" isn't clarity...it's a kind of burnout talking. Which is, of course, understandable. You're allowed to feel that.

But I don't think it's fair or illuminating to erase the very real ways this stuff has made life better for billions. It's not perfect. But better. And worth building on.

xwowsersx··on AI makes the humanities more important, but also weirder
I have no doubt AI will impact the field, as it will many (most) others. That said, some assertions that may warrant reconsideration IMO:

> The machines best us across nearly every subject

This feels overstated, given the well-known limitations of AI in reasoning, factual reliability, and depth in specialized domains

Describing traditional scholarship as mere "archaeological artifacts" feels like prematurely dismissing the enduring value of books and other research practices

- The idea that AI offers "pure attention" seems to romanticize what is, in essence, statistical pattern recognition...potentially misrepresenting the nature of the exchange

> Factory-style scholarly productivity was never the essence of the humanities

maybe inadvertent, but devalues serious academic work that follows rigorous, methodical research traditions

Beyond that, there's a few claims that would benefit from some additional evidence:

- student paralysis rely heavily on anecdote; broader empirical data across varied institutions would strengthen the case - His comparisons between AI and human experts feel selective...would be interested to see a more systematic evaluation to see where AI meaningfully matches or falls short of expert performance

- I'm somewhat partial to the author's view, but in terms of long-term educational outcomes, no real evidence to show that AI integration improves learning or contributes meaningfully to human development - the idea of a "singularity of self-consciousness" is pretty... under-argued? - The supposed link between declining humanities enrollment and AI's potential lacks concrete data to support the correlation

xwowsersx··on The Polymarket users betting on when Jesus will return
Story of my life, my friend
xwowsersx··on The Polymarket users betting on when Jesus will return
fare* poorly
xwowsersx··on Why walking is the most underrated form of exercise (2017)
I train six days a week, but since I work from home, I was actually pretty sedentary outside of my gym time—basically just an hour or so of movement a day. That changed recently when I got a Fitbit. I've gotta say, it's helped a lot. Just having something simple like step tracking and reminders to move makes it way easier to stay aware and hold myself accountable. Now I'm consistently hitting at least 10,000 steps a day. Before I started tracking, I would've assumed 10,000 steps was a lot, but now I realize it's actually pretty easy to hit if you’re just a little intentional about it...tbh 10K feels like a basic minimum.
xwowsersx··on Show HN: Chat with 19 years of HN
Should anyone from camelAI be present, a quick note: the page at https://camelai.com/data-sources currently renders blank—none of the data sources appear.

EDIT: oh, I guess that'd be you, vercantez :)

xwowsersx··on Pyrefly: A new type checker and IDE experience for Python
Happy to see instructions for integrating into Vim/Neovim: https://pyrefly.org/en/docs/IDE/#other-editors
xwowsersx··on Thoughts on thinking
Well said, and such an important point.

> Developing a prompt is like scrolling Netflix, and reading the output is like watching a TV show.

That line really hits home for me.

xwowsersx··on Perverse incentives of vibe coding
It's unlimited? I thought it was "Choose 5x or 20x more usage than Pro." Could you please elaborate?
xwowsersx··on Void: Open-source Cursor alternative
Can you use OpenRouter with this?
xwowsersx··on Replacing Kubernetes with systemd (2024)
FYI the link to Flux in

> Particularly with GitOps and [Flux](https://www.weave.works/oss/flux/?ref=blog.yaakov.online), making changes was a breeze.

appears to be broken.

EDIT: oh, I hadn't realized the article was a year old.

xwowsersx··on Show HN: GitMCP is an automatic MCP server for every GitHub repo
For those unfamiliar, this is similar to taking a codebase and processing it with a tool like gitingest, which transforms the entire repository into a format suitable for LLMs, eabling contextual convos and queries about the code. The additional component here is the integration of the MCP protocol, allowing any compliant model to interact with the provided MCP server and dynamically query the codebase to answer questions in real time.

Not sold on MCP being the right paradigm (we'll see), but had a lot of fun building an MCP server recently using https://github.com/tadata-org/fastapi_mcp to quickly get up and running and be able to call to it from Cursor.

← PreviousPage 3 of 34Next →