jonahss

I love how this article reads similarly to articles about recent advances in the Conway's Game of Life community.

From my point of view: a team combined all the tricks in the community into one machine and we finally created a new life form to play with!

I feel kind of good knowing that my code, design decisions, styles, are now part of the data shaping all software now.

It overhauled my media server.

I hadn't set it back up after moving. I gave OpenClaw ssh credentials and it updated the OS and packages, then couldn't get back in after a restart.

I plugged in keyboard and screen and it was stuck at boot, couldn't mount a drive.

I sent OpenClaw screenshots and it told me to type in journalctl commands. Then it had me modify fstab so boot could continue.

After that, OpenClaw could get back in on its own. It found the drive I'd been using had 1300 bad sectors and was going to die. It saw that another drive was perfectly healthy. It said the bad disc sectors were all early and probably just filesystem metadata and my files were probably fine. It copied 1.5Tb to the newer drive and restored everything.

I probably would have thrown the whole box out, as I hadn't used it in a year and wasn't looking for a project like that.

It seemed to me that pre-llm, google had stopped surfacing stackoverflow answers in search results.

One thing that China has, that the US does not is a marketplace like Alibaba. It's simply incredible how you can place an order for any component with a quantity slider that goes from "three" to "three full train-cars a day".

When starting a hardware product, it's so easy to buy small quantities at commercial prices, no questions asked.

In the US, it seems all sales is relationship-based. I need to fill out forms on bad websites, wait for emails from sales reps, who sniff out immediately that I only want a small order. The overhead for processing an order isn't deemed worth the time, and any US-based small-time project ends there.

If anyone is interested, I know someone in San Francisco who built a full injection molding shop in their basement and is happy to take orders.

Also I know a 3d printer that can print molds for injection molding, and I know a guy with a set of printers who'd be happy to take orders.

I think the best part was the little video-game video of Stevens checking different datasets by walking around. Love it.

These companies are not trying to be companies that sell an LLM to summarize text or write emails. They're trying to make a full Artificial General Intelligence. The LLMs pull in some money today, but are just a step towards what they're actually trying to build. If they can build such a thing (which may or may not be possible, or may not happen soon), then they can immediately use it to make itself better. At this point they don't need nearly as many people working for them, and can begin building products or making money or making scientific discoveries in any field they choose. In which case, they're in essence, the last company to ever exist, and are building the last product we'll ever need (or the first instance of the last product we'll ever produce). And that's why investors think they're worth so much money.

some ppl don't believe this cus it seems crazy.

anyways, yes they're trying to make their own chips to not be beholden to nvidia, and are investing in other chip startups. And at the same time, nvidia is thinking that if they can make an AI, why should they ever even sell their chips, and so they're working on that too.

No way, the real Yu-gi-oh is played with hologram projectors and the cards have personalities. Plus the rules aren't ever quite explained.

I've heard that switching to all of that will add 20-40 seconds of dead time waiting for the display to change, as the NFC transfers power to run the whole procedure. That'd be too long an interaction time with no feedback.

And of course the cost goes way up.

> It just seems that from every angle I look at this thing, I see clear problems with bringing it to market as a viable product.

Yeah, the goal is to make a cool thing, and then make a fun game.

So, the goal was to build a devkit so that I could use it to develop my own game. I wanted the physical things to exist, so that I could try it out with people and find out which interactions are fun, vs which aren't.

Well the internet liked the idea, and saw some of the same promise in it that I did (plus HN is a sucker for e-ink). With all the interest, and people asking how they can get their hands on one, I ran a crowdfunding campaign to make devkits for everyone who wanted.

Turns out making 25 of something is way more work than making 2 of a thing. The supplier changed the display firmware on me, I had to make things to more exact measurements so parts were interchangeable, I had to write docs and make videos, etc etc. Took a whole year.

Now the pressure is off and I don't owe people products after they gave me money for them, so I can take a break to clear out my backlog of minor projects, then get onto designing my own game, using my own devkit :)

Once I have an actually fun game, I could increase volume and bring down the costs. My goal is to make the game accessible for $80. I'll need the e-ink price to come down a little, and use injection molding instead of resin casting and wood. Plus the base won't need a Raspberry Pi, that's just for my quick iteration. Final product will need to be embedded.

(Almost all the devkits sold were to friends and family, with only a few going to actual game designers. CrowdSupply itself puts in an order for more units along with the campaign, so that they can stock them after the campaign ends and initial delivery is over. Except they negotiate a different price for those because the margins work differently)

I didn't mean people who work at Uber explaining how their load-balancing works.

I meant walking up to a super cool music visualization at an outdoor art festival and they guy there happily explaining to me their entire system built out of a node flow diagram implemented on Max but adapted to visual graphics using a plugin called Vsynth.

Or going over to a friend's house and seeing their modular synth system and they happily explain to you how it works for an hour.

Just this weekend I met an amazing engineer with a street-legal steam-powered motorcycle which he patiently explained for an hour.

The pcb on the card is just a charge pump that the display demands in order to regulate its voltage. If I shipped that off the card and onto the base, I'd need more contacts.

Those flat flex displays are awesome, though way pricier. Then I'd lose the nice stiff pcb board which allows all the contacts on the back to mate. I could go wireless, but then the power delivery needs to charge a capacitor and the delay between button press and display refresh will go to 30 seconds or beyond

The lightnote is so cool, and the design is perfection

I've done a little bit of that, but not as much as I should. I took it to a boardgame convention and a GDC event, but not much came from it. If anyone has introductions or advice there, I'd appreciate it.

Haha, guilty. But! The actual wires assigned to each pin correspond to the Sephirot. Keter is VCC and Malkhut is GND, etc etc :P

I'm hoping to monetize the Shadow Realm

:wave:

Yes, still no game, I spent the last year delivering the crowdfunded devkits.

Ah yeah. TCG is just a quick way to explain it to people so they get what I'm going for. I personally dislike the whole "rarity" aspect of TCGs, though it was fun when I was a kid. Living Card Games like Netrunner are more fun to play IMO.

My elevator pitch usually goes something like: MTG meets Yu-gi-oh meets Pokemon (the RPG) meets Tamagotchi.

Wyldcard creator here, thanks to HowToWare for the interview.

Since the intention is a children's toy, I'm trying to get the price as low as possible. This display size is cheap because it's intended for grocery store price tags ;)

Flat Flex e-paper displays are available but much more expensive. Plus, the magnets in the cards which let them mate to the base take up most of the thickness. I'd have to rethink that physical interface entirely if they needed to be much thinner. My next step will be to design an actual game and then I can come back to size for a v2.

Wyldcard creator here, thanks to HowToWare for the interview.

It's a link cable to connect your base to your friend's when you play against them. I link it into itself just to keep it tidy and it doubles as a strap to carry it by.

Hi, Wyldcard creator here, thanks to HowToWare for the interview :)

The cards can store data, and I imagine that things which happen to the card during a game leave a lasting impact, which is carried from game to game. The cards grow and change over time, and so when you trade one, you're trading an entire legacy :D

In 2018 I managed a team at a marijuana startup and was evaluating Triplebyte for recruiting. One evening, after sampling our own product (post-deploy smoke tests), I went through the Triplebyte quiz to see what my candidates would experience.

A couple years later I was job hunting myself, and went back to my Triplebyte profile. I sheepishly admitted that I didn't bring my A-game to the quiz, but was told I couldn't retake it, even though it had been a few years. Odd to me that there wasn't a provision to account for candidate's skills to grow over time. Got plenty of interviews anyways >.>

I used Triplebyte successfully for recruiting as a manager at two companies, and was sad to watch it decay and die while working at a third :(

I hired some of the best people I've ever worked with on Triplebyte.

sounds neat. link?

I agree. And I think the key to this is VR.

Another comment in this thread is about how hard graphs are to visualize, but a 3D interface gives you a lot more room.

When VR hype began I thought "well what's the excel of VR?". Microsoft's answer was "2D spreadsheets floating in 3D space". What nonsense. I think graphs.

email my username at gmail.com if anyone is interested in exploring this together!

There's another important element: the ability to judge quality.

Many people (especially baby boomers) learned a shortcut, by using brand name as a proxy for quality of goods. The problem with this is that all those brands eventually outsourced and sold themselves and cashed in on the old brand quality association. Cheapening and cheapening as the producers realized that people still bought their product no matter what the quality was. Also, people equate flair and style of product with the quality, because they aren't actually good judges of quality. You can find many cheap junk products today that retain a poor skeuomorphic shadow of their former glory. At some point consumers learned "well if it has the shiny chrome it's a good one" and producers learned they could add a chunk of shiny plastic and people would prefer their product.

Brand names are worthless now, for the most part, and if you're not good at judging quality yourself, it can be a difficult consumer landscape to navigate.

I agree. I feel like the author is upset at pens given out at banks and restaurants to sign stuff. As cheap as possible, Minimum Viable Pen for sure.

I buy a few packs of my favorite uniball pens every year and sprinkle them around the house and in all my bags. Very happy with them.

one thing about some of the shitty products out there today is we got MUCH better at making things out of less material. So they break easier and wear out faster, but we've been able to reduce material usage along with price.

If you're a person who loses pens or holds onto them so long the ink dries up, it'd be better to waste an object made as cheaply as possible than one made to last.

This is a balance to play as well.

the Miele vacuums are incredible.

If you want appliances that last, buy the commercial ones. Hobart makes an under counter dishwasher that you can bet will never wear out and can be repaired.

Shell out for an industrial washing machine like they have at the laundromats. Those things are made to run all day for years. If they break down, they're repairable.