←
Luke Toledo2,323 posts
Luke Toledo avatar

Luke Toledo

@lukeSVG

intersectionist, designer

◷ Joined April 2021▧ 309 Photos and videos
Luke Toledo avatar
Luke Toledo@lukeSVG·1h

Made a timer-like device with Astra 6.2 + Blender MCP since its meant to be great for 3D. Really like the results so far, haven't tried for coding or other uses yet. Used Sol 6.1 to discuss the idea, @schematikio to get the component layout, then to Astra 6.2 for the realistic draft + exploded version. Also tried Sol 6.1 for the model, and you can see the comparison output. For the rotation video, I'm using Remotion + Sol, since its a simpler view. Happy with the results. Stuff you dreamt of doing decades ago but could never justify, now simple hobbies. It's amazing how much creativity it enables.

Post attachment
0:05
Luke Toledo avatar
Luke Toledo@lukeSVG·1h

I've done cash-based contests before... (good and terrible) Higgsfield et al needed to cut thru the noise, so they lead with a big number: $100,000.000 DOLLARS. It captures the headlines, but the lack of tiers (by audience in this case) means this is is now imbalanced. @indexsy has a 25k sub newsletter (I'm there) so he leverages that—but that means its one and done. time will tell, but I suspect this will now mostly brew negative sentiment, and the attention is capped. Shoot-from-the-hip approach can work in indie dev, but not in cash prizes, because perception matters. For what its worth, I think everything is done with the best intentions.

day 3 of the $100K hackathon. 4,800+ builders, someone's already at $9,600, and 28 days left to knock them off the throne 👑 free to join, link below

1:04
Luke Toledo avatar
Luke Toledo@lukeSVG·Nov 17

The risk of creating AI content is that it adds to the same noise that the business is trying to cut through. Its a “tragedy of the commons” type situation where new businesses kill their own marketing channels. For example, you know AI-built SEO articles are bullshit, but you feel that you still HAVE to do it, otherwise your business drowns. But because there’s so much noise now (socials included), we’re at a point where if you want to be found, only one thing will work in the future: someone HAS to recommend you, either a trusted person or agents. And trust will be the most valuable commodity, not distribution (which is just reach). Might be worth start thinking in trust, not just eyeballs.

Post attachment
Luke Toledo avatar
Luke Toledo@lukeSVG·Nov 16

The risk of working with LLMs is how addictive they are, and they feel more like gacha games than another productivity tool. And because it triggers our dopamine reward pathways so well, it drains us at the end of the day. You get results that "almost work", it sucks you into "just one more prompt", and suddenly you're in a Claude-hole. I have too many people around me describe anxiety and exhaustion after long AI sessions. The gacha similarities are there: - You submit a prompt, wait, and find out what you get. - Rewards are variable, and there is a component of skill. - Every "pull" costs something, token or attention. - You can re-roll your system over and over, new characters to get different rewards (which i think often expresses in "software factory" rabbitholes) - But there's leveling! The more you build, the bigger the rewards and outcomes. However, the best scenario seems to be mixing LLM with slow work—pen and paper, walking and thinking, stuff like that. I think it helps perform better over-time, the equivalent of a active mental holiday.

Post attachmentPost attachment
Luke Toledo avatar
Luke Toledo@lukeSVG·Nov 12

Claude is clearly going after enterprise, and they need to, if they want to have any chance at a successful post-IPO price action. So indie devs get left alone. Plenty of challenges still, but at least this won't be one of them. Something else interesting is how this supports collaborative ownership, which imo is probably the last "moat"... your own ability to choose and own labour outcomes.

Introducing Claude Meetings: add Claude to your team’s video call and it builds docs, whiteboards, and other artifacts with you while you’re still talking. Starting to roll out in beta on Team and Enterprise plans.

1:03
Luke Toledo avatar
Luke Toledo@lukeSVG·Nov 12

every morning kickstarting Claude

Luke Toledo avatar
Luke Toledo@lukeSVG·Nov 8

wtf this is so good

i'm a solo dev and i built an infinitely evolving backrooms game and released it on the web. 35,000 players and ~9,600 hours played. these are my metrics under 36 hours of launch. each blue dot below is a player on the map i recorded a moment ago. there's ~1,200 players playing the game at THIS moment. SIMULTANEOUSLY. let that sink in. this is incredible. we're entering a new era for sure. note for video: only shows 200 players in total because my server chunks the shared space into worlds of 200, so 1200 players would be 6 different worlds and you're seeing one of em. i'm building out a shared space server that can sustain 2000 players in a single world. launching real soon.

Post attachment
0:21
Luke Toledo avatar
Luke Toledo@lukeSVG·Nov 8

Anyone can be a Design Director if they want to these days, but now to agents rather than people... and its kinda incredible I used to work as Design Lead to mid-sized startups before, and my mornings were often lots of review requests in the morning, plans and meetings in the arvo, bit of "pencil" work by end of day. And now I basically do the same, but with agents: 1. Sol screenshots every screen & state in your app 2. Lay it all out on Figma or Paper 3. Create a "Coordinator" - you're going to ramble 4. Ramble! Voice review things to your Coordinator, and its job is to make sense of things and create new sessions ("Designers") 5. review Designer individual work, and once done, approve and send to an "Implementer" - its job is going to be coordinating the work into your branches Have a design.md + components and 80% will be done per spec Keep feeding open source repos as references per component, so you don't have to waste time fixing interactions Will you get S-tier design level craft? No. I think thats still the domain of human designers, because every component and animation has endless edge cases to work right and feel right. But will you get far better output than AI slop, and do the work of 20 people teams as one? yeah... And you just become responsible to juggle the chaos in your head, make sure UX/UI remains coherent, and you are directionally correct in your product vision I frankly get kinda scared for juniors these days, these used to be the ropes to learn, and now they are going away

Post attachmentPost attachment
Luke Toledo avatar
Luke Toledo@lukeSVG·Nov 4

why do newsletters think we're buddies? I think its a major fuckup, and most send re-activations like I remember you are at all... this email is just one of many, just goes straight to trash. remind me WHY I should care again. Most software now normalizes adding ANY sign up to a list, which is fair because we all want to be able to activate churns, but... I don't remember who you are! Keep reminding. Tell who you are again, and why it matters. Maybe put some context, for your own CR sake?

Post attachment
Luke Toledo avatar
Luke Toledo@lukeSVG·Oct 30

bought into the @paper hype because designer twitter would not stfu about it it doesn't have components. not even a little work around. are all these hype people imbeciles? i don't get it