Vince Buyssens

bruteclauding a semantic design system for my brand world with cursor ► thoughtform: navigating language and strategy ► atlas: navigating worldbuilding & aesthetics ► astrolabe: navigating knowledge and keynotes It all started with the insight that "AI sees meaning a geometry you can navigate through", which I've been trying to encode into design primitives and tokens starting with a grid system inspired by navigational interfaces now building a design navigation system that uses embeddings to match design references to my existing components i've been obsessed with design and interfaces my entire adult life, but never had the skillset to do anything with the fact I've already gotten this far without any coding experience or design background is insanely validating, but also speaks to how good models have become (especially Opus 4.5) makes the $100 i almost spend per day on the ultra plan worth it (but maybe Santa can sprinkle some credits through my chimney @ericzakariasson 😇)

@mooncat_is "before someone takes this down" sure my dude

It turns out nobody at Anthropic actually knows how to build an LLM, we’ve just been asking an old copy of Bing Sidney That means if you can _just_ repeat this lecture word for word (before it gets taken down definitely soon), we’d be immediately forced to give you all our money

@hecubian_devil yes! really like how Michael Levin thinks about this: intelligence as a continuum with AI as part of it

I don’t think LLMs are conscious and I doubt they are structurally capable of conscious experience. I’m partial to the idea that embodiment is a precondition of consciousness. But, full cards on the table, it is a little annoying the absolute *certainty* with which the anti-AI side (broadly, my side of these debates, I guess) proclaims this, given that we basically don’t understand what consciousness is, how it works, what gives rise to it, etc, in human or animal brains. Scientists can’t even agree if plants have subjective experience. Like, if you can’t empirically detect consciousness, explain its workings, understand what physical processes give rise to it, or even know which living things experience it, it seems to me you must have at least a *little bit* of epistemic humility about the whole thing. If we actually understood human/animal/plant(??) consciousness well, that would be one thing. But we don’t. We notoriously know almost nothing about it. Now, this would be annoying but basically trivial *if* we weren’t also ceding an opportunity to leverage the belief some people have in AI consciousness as part of mustering a broad political coalition to regulate AI. But it is really, really annoying that we are neglecting one possible avenue for mustering that support out of a totally irrational and empirically indefensible *certainty* about something which science and philosophy both fail to understand at minimum levels of adequacy (that something being human consciousness). Like, we’re throwing away a tool for reasons of pure ego, essentially. Again, I don’t think the machines are conscious. I strongly doubt LLMs are architecturally capable of *ever* being conscious. But I wouldn’t stake anything of serious value (like an organizing opportunity) on my belief here because that’s a waste, and *also* I really can’t claim certainty in my belief when I have no way of empirically evaluating whether something is conscious or not! Because none of us can!

@shadcn yezzur

Perfect.

@poellll just love how the Destiny's worldbuilding could support both the gothic aesthetic of a Thorn and the slicker futurism of a Graviton Lance

while i agree to some point, destiny was borderline gothic fantasy disguised as baroque futurism and therefore very valid in itself

@levie i disagree; even if dawkins is wrong and AI never achieve 'consciousness', AI is the first technology that sits between a tool and collaborator it's not like us, but you really can't treat this like normal software tbh

In general, we should treat AI like a utility, not like a being. The more we confuse what AI is the more we will make ourselves go crazy with analogies that will never fully hold true.

Not exactly what you're looking for @TaylorLorenz, but here at Loop Earplugs we've actually decided to not do AI video anymore even though we made two pretty good, successful, non-slop ones We use AI in paid social for stock photo replacements and product shots, but AI for above-the-line just isn't worth it and people don't really want it either x.com/voidwalker_com…

@Stunt_ManMik3 thanks really really cool! Reminds me a bit of The Dark Crystal: Age of Resistance where they beautifully blended puppetry, practicals and CGI my favorite VFX is probably the subtle facial expressions they mapped on the puppets to make them more emotive

@voidwalker_com So its a set shot blended in to a miniature with a cg background and its kinda haphazardly done but ends up looking awesome. Its more apparent on a tv not recorded on my cell phone lol

Man, there’s something super endearing and awesome about movies from the 90s having clunky effects work, but the look is so stylized & there’s a clear visual identity that it works and enhances the vibe of the movie. Soldier Mortal Kombat Alien3 Etc

@rymagno @contralabs_ai will you be offering this headless? Would love to test this on our creative briefs / ads at Loop Earplugs!

AI is closing fast on best practices—prompt adherence, typography, CTA placement. Taste is still ours. The convergence vs divergence finding is the one I'll be quoting all year. Proud to be part of this @contralabs_ai.

@irace @stripe @jeff_weinstein @jrfarr @kevinyien @hazelcough @BackseatVC Done!

@voidwalker_com @stripe @jeff_weinstein @jrfarr @kevinyien @hazelcough @BackseatVC Yeah please — irace@stripe.com

Our roadmap is now public: see hundreds of upcoming product ships. We’ll be adding to this list regularly. What’s missing? @jeff_weinstein @jrfarr @kevinyien @hazelcough @backseatVC and others will be answering your questions today. 🧵

@irace @stripe @jeff_weinstein @jrfarr @kevinyien @hazelcough @BackseatVC feel free to dm!

@irace @stripe @jeff_weinstein @jrfarr @kevinyien @hazelcough @BackseatVC would you mind if i send a concise email with the business case / frustration / solution? i know you guys are super busy, but the fact i have to download all these invoices manually really kills my energy every month; while i think this challenge is super tractable

@irace @stripe @jeff_weinstein @jrfarr @kevinyien @hazelcough @BackseatVC put way too much time into this mockup, but I really think people would LOVE this

@voidwalker_com @stripe @jeff_weinstein @jrfarr @kevinyien @hazelcough @BackseatVC Fully agreed, thank you for the feedback!

Ah yes good to know, but indeed doesn't change the spirit of his observation. Genuinely looking forward to your book, because despite its downsides i really believe AI rewards people who can think across domains The richer your frame of reference, the more AI can make connections across it.

In 1st printing of Range, I put that quote in the context of a "high school graduation speech" that Graham wrote. It was a speech, but not a graduation speech. He pointed out the error; I corrected it for all future versions. Does't change the quote, but important to acknowledge corrections!

Holy cow, that is crazy praise coming from a thinker I admire 🙏

i also found your insights to be the biggest unblocker when it comes to working with AI. i was listening to a podcast with @benedictevans the other day and really like how he framed that the biggest challenge with AI is "working out how to ask for what you want" which your books answers beautifully in so many ways. i quote this quote from @paulg "Instead of working back from a goal, work forward from promising situations." regularly

@voidwalker_com That is amazing to hear, Vince. Thanks so much for the kind words.

@DanDavisWrites best voice in the industry

What strikes me about Crowe is how he defined a kind of deeply moral masculinity for a generation in the 2000s (say 97 - 2010). Every big role he chose and crafted portrayed men grappling with duty, honour, responsibility. We all love Master and Commander but the Insider, Cinderella Man, LA Confidential, 3: 10 to Yuma, Robin Hood, the Next Three Days, etc. They're all about men grappling with the consequences of doing what must be done. Legendary run, dunno how he pulled that off.

big believer in software-for-few, like here at Loop Earplugs I built a pipeline that can transcribe-translate-dub UGC videos at scale. Localization is a big revenue lever for us, and this allows us to bring our best-converting format to more markets (all with the creators' permissions of course). all built with zero programming background; i'd say my background in journalism has been my biggest force multiplier because it taught me how to best articulate my ask, and how to ask the right questions.

Agentic coding is a huge boon for software developers that want to get far more done, great for IT people to build vastly more custom systems internally, great for domain experts that want to automate workflows or wire systems together, and absolutely fantastic for anyone curious to learn how to start coding. What it’s less great for is casually building complex software that you have to maintain on an ongoing basis and take on all the risk for. Upgrades, maintenance, keeping up to date with latest security issues, and so on, are taxes most knowledge workers aren’t familiar with or prepared for. Net net: we’re going to get 100X more software and vastly more software developers in the future. But that’s different from *everyone* rolling their own.

@zeynep interesting anecdote on this: here a Loop Earplugs Claude skills work really well for paid social / CRM copy (content we have hundreds of pages on); I would say it almost one-shot in, but when it comes to my personal prose I always have to do a lot of editing

While this may work on some AI detectors, the method misunderstands the key differences between AI and human writing and will falsely tag some good writing that LLMs emulate. “Generative AI is extremely impressive but not on a path to AGI” repeated daily can help illuminate.