elicash

This was fun!

Neither of us have any idea whether his manufacturing startup is successful. It looks to be a 5 person team. A google search suggests they had a funding round early this year.

It might, in fact, be that you are more successful with your approach. I have no idea. Congrats if so.

That's the cost of one employee (and not an expensive one). Do you think it's impossible he's getting more value from the AI than he would from a single employee?

Yes, this dissonance is the point of the post. The first two sentences use the word, and then it's repeated throughout.

It's nice to see a person who actually acknowledges tradeoff.

For the curious, this seems to be his company: https://www.umh.app

The non-branded version from Work Louder seems to cost $174 https://worklouder.cc/creator-micro-2

So I'd say it's probably only a $56 contribution.

RE: the joystick. They should have gone with the Playdate's crank instead, CrankGPT style.

Anyway, I think it's all a fun marketing thing. A desk toy for folks with disposable income. I imagine they'll sell out, given the limited release and then they'll be on eBay.

What do you plan to do next with it?

I'm looking for good and cheap transcription + speaker diarization on my Mac for a small personal project. Recs?

I'm not sure what source would make more sense here. This person cited a twitter account, but that person didn't ever post it. So this is the source, best I can tell.

This is a person with OpenAI labs, which created it. They cited their coworker who didn't post it themselves.

Source: https://x.com/keyanzhang/status/2075374891558425037

It amusing that Musk attempted to reverse his purchase of Twitter by citing the number of bots, and then research like this comes out alleging that now 29% of the X's long form articles are fully AI.

It's not exactly the same thing, of course, but still interesting the extent to which this type of content is viewed as the business opportunity for him.

I think if you occasionally tie gym shorts and don't have needs around fit, this is correct.

But if you're a nurse and wear scrubs all the time, then it might become worth spending time upfront perfecting it.

This is the attorney who runs https://nlrbresearch.com (which is run on Datasette) and the AI assistant (Claude Skill) that is tied to it https://www.youtube.com/watch?v=mOa3Pc8RD40

This isn't true. He said "And other than when I was 12 years old I have never cheated in a tournament with prize money." That implies, to me, that when he was 12 he did in fact cheat for prize money.

And then he admitted that he cheated as a 16 year old in "random games" to boost his online rating. I can't find any instance of him providing a number of how many games he admits to cheating in, but the chess.com report alleged over 100, including when he was 17.

This is like how I can't tell the difference between good wine and mediocre wine. It's not a skill I have a desire to develop.

I don't agree with the overall point, but what they are saying is that if Tidal makes a "lofi hip hop mix beats" playlist, then they have to pay a nickel to an artist if they direct the user to their song, but NOTHING if they direct to someone who uploaded an AI song. So that's an extra nickel of profit for pointing to the AI song.

In reality, I don't think this is how it'll shake out. But it's a valid argument.

If there's a market, then a competitor app with different policies will likely arise.

The other aspect that's missing from the discussion here is LEGAL. If Tidal is making money from stolen music -- although arguably they still are by offering it on a subscription basis -- then that opens them up to litigation. From that perspective, this may double both as risk-mitigation and also a marketing opportunity for them, would love an attorney to weigh in there.

(From the comments here, Spotify is the market leader and already pays out for AI generated music. But I can't say that independently.)

> Tidal defines AI-generated music as music that is wholly or substantially generated by generative artificial intelligence.

I think this needs more clarity. I can think of a lot of different ways AI is used in music today as a part of the song generation process and not sure whether or not this definition would apply to it. They specifically mention developments in "text-prompted generation" but if anything that confuses the issue more, for example what about training on specific music.

This isn't a comment on how expansive or narrow the definition should be, just that they need to spell it out more to allow for consistent application (to say nothing of enforcement). If someone uses ChatGPT for lyrics, but writes the instrumentals themselves, does this policy apply? I genuinely have no idea.

I think the Neo was eating into their Air sales, and not merely bringing the Mac to a new market.

I think when they eventually announce the M5 Mac Mini (September?) it'll just be at a higher price.

A fix so amusing and ridiculous it gets the bug enough attention that Apple fixes it for real.

> Have a play-by-play view so users can see plays they missed

I'd like this not as a timeline scrub but as a back button.

Great! Since you're taking feature requests, I'll add that it seems like sometimes lots of data comes at once and then everything gets rushed. But then nothing happens for a while. So it has a "hurry up and wait" feel.

I think it'd be worth slowing down when you have receive data for multiple pitches at once, which would also mean it wouldn't feel as slowed down at other times -- even if that means at some points you're further behind the real-time nature of the game.

Could you display live data in a browser tab title? Not just the score, but actually mostly when it's not between innings so I know when to switch back to it.

I also wish it was more zoomed-in on the action, but I get why that's not in v1.

Could you upload a video?

You've posted this from multiple accounts

It'd be $8.52 in 2015 dollars, but certainly they are the ones who mentioned the $12 amount not you, so I'll put that aside.

Far more importantly, you would not get billed for 2 minutes of work for this if you paid a developer to fix it. At best, half hour increments for the fix. But more likely, for the full hour. Also, in this comparison, the consultant is on call every day, morning, afternoon, evening, for whatever you wanted and will jump on the job immediately.

I don't think it's that simple. (I generally agree with you; I just that that oversimplifies.)

Another model might have used fewer tokens, but come up with a fix that was 1000 lines when the right fix was only 2 lines.

I tried using this calculator: https://www.andymasley.com/visuals/ai-prompt-footprint/

It doesn't have Claude Fable yet, so I went with GPT 5.5 Pro. And so I'd estimate it at 22 gallons of water used (different from consumed, of course). That's quite a lot! It amazes me how much the different use cases and models use dramatically different amounts of water. My takeaway from playing with that calculator has been the folks who talk about water usage are overstating the impact of chatbots, but not overstating when it comes to vibecoding.

The good thing is that competition should drive down how efficient these models are in the long run. This blog post makes me not want to run Fable because of the cost, and that incidentally also means selecting models that aren't as wasteful in terms of water and electricity.

I misread your comment at first and thought you were insulting Simon Willison, rather than calling Claude Fable a bad developer, and so I'm commenting here to clarify it in case others also misread it.

That first sentence threw me off.

Anyway, I'm glad he spent the $12 because this blog post was highly informative.