darrinm

App stores are approving an order of magnitude increase of new apps because people are now able to create them using natural language. End users are deriving great value from them. So this is settled?

How does Muse stack up against Grok Bot capability/feature-wise?

Shouldn't the biggest concern be that OpenAI either doesn't know about these breaches or is concealing their knowledge of them? I mean, as of yesterday their primary message on this track is "most aligned model yet".

Are they doing this as a means to detect distilling? What kind of encoding would survive that process?

A simple prompt that still stumps frontier LLMs most of the time is “create a pinball game”. They’ll put all the right pieces there but then fail to arrange them such that the game is truly playable. They’ll put a wall in the way of the launch chute so the ball can’t be launched. Or the flippers will pivot the wrong way. Or there will be holes such that the ball drops off the bottom without getting within reach of the flippers, etc.

Opus 5 is the first I’ve seen to “one shot” it (in a harness, so it was more than one LLM call).

Yeah, /model fable also worked for me (despite not being shown on the /model list). Thanks.

Not supported in Claude Code yet?

As a kid I didn’t understand what the 555 timer chip on the Apple II disk controller was doing but I learned the hard way that when you misalign the pins on the drive connector cable and the 555 chip releases its blue smoke you can’t use the drive anymore :(

They announced back in September that they’re coming to Seattle. https://www.king5.com/article/traffic/traffic-news/autonomou...

I hear it a lot but this gambling analogy breaks when you look at actual outcomes. If you went to Vegas and after a few pulls on a one-armed bandit could _reliably_ walk away with the jackpot we wouldn’t even call it gambling anymore.

FWIW many tool calls can be and often are made in one inference cycle.

Has anyone tested how generation speed compares to gpt-image-1?

Did you try the strict (beta) function calling? https://api-docs.deepseek.com/guides/function_calling

Amazing write up as well as the other implementations and writeups by the same author it links to.

More relevant is if by _playing_ it they couldn’t tell which is which.

It confused me that their stated evaluations by humans are comparing video clips rather than evaluating game play.

So… is it interactive? Playable? Or just generating a video of gameplay?

Shameless plug for my startup’s freehand web editor https://hatch.one. You don’t have to throw back to hotglue if you want to create your own site visually with tremendous creative freedom.

Hatch (https://hatch.one) has many of the characteristics of HyperCard plus its "projects" are full featured, shareable and remixable (if you want) web pages.

- Easy drag-and-drop interface with a library of images, audio, video, etc.

- Programmable with Visual Scripting for an easy on ramp to adding logic and behaviors.

- Programmable with Javascript if that's what you want to learn.

- HTML "components" if you want to dig into HTML, CSS and all that good stuff.

- A few tutorials to help get started with physics, animation, etc.

- Free.

- Super fun! Ok, I'm highly biased as one of its creators.

This article has been such an inspiration for us at Hatch (https://hatch.one/)! We founded the company as “Personal Software” and we’re working hard to lower the barriers for this kind of creation. The opportunity shouldn't be limited to people who know how to code. Several pieces of the puzzle are in place with more in the pipleine. Here's quick video of getting started creating a web app in 60 seconds: https://www.youtube.com/watch?v=HQMFFkCHrdo

Love rough.js. Thanks Preet Shihn! I used it to create Sketchy Shapes you can add to Hatch projects. Fun to play with the parameters in realtime: https://hatch.one/@darrin/sketchy-shapes/edit

Benchmarks?

Should clickbaity headlines be banned?

Note that, unlike Chrome, Safari only supports “origin private” files which means it cannot be used to access files already on your system (e.g. no vscode.dev, photoshop.adobe.com opening of local files).

Everybody wants a different set of App Store rules. The problem is that there is only one App Store and one company choosing the rules. With competing App Stores (e.g. “Great for Kids Store”, “Adults Only Store”) we’d have freedom of choice instead of single corporate rule.

Imagine if one company set the rules for allowable websites. What would (wouldn’t) they ban?

Instead the Web is wonderfully open and parents, schools, individuals, even nations can choose configurable filters to have a web that aligns with their values.

The Roblox iOS app is entirely populated with non-Apple approved user created games. It also streams/downloads (non-Apple reviewed) code on demand. This seems to be against a number of App Store policies.

Maybe Apple is OK with all of this when they get their 30% cut as they do with Roblox because it uses Apple's in-app purchases?

Why is it ok for native apps to do all such (and more) fingerprinting, utilized across apps, and not the web? Maybe I’m saying this backwards, why aren’t native apps given the same restrictions?

Agree. If coughing into my arm helps (CDC recommended) it’s hard to imagine any kind of mask wouldn’t help too. Also recommends mask: https://www.cdc.gov/flu/pdf/protect/cdc_cough.pdf

How will this affect caniuse.com? I use it every day.

Sounds like you’re describing the Progressive Web App (PWA) install that Chrome recently added. Coming to Firefox (if not there already), and Edge. Not holding my breath for Safari though.