Jonathan Fly 👾 @jonathanfly

CEO of bad ideas. Using the wrong tool. The least efficient way. For no good reason. 👾 https://t.co/9O94rxu31k https://t.co/DDPn3Nlhom github.com/JonathanFly/ Joined April 2009

Tweets

7K
Followers

6K
Following

3K
Likes

43K

Jonathan Fly 👾 @jonathanfly

2 weeks ago

@1thousandfaces_ @aliceisplaying Was the old CoT always this coherent and natural sounding? I've also never gotten anything quite like this, but probably didn't see a lot until 5.2> either.

0 0 0 5 0

View Details

Jonathan Fly 👾 @jonathanfly

2 months ago

@emollick Love that first prompt

0 0 1 264 0

View Details

Not uncommon as a game mechanic (I understand the tweet is a joke about the real world) for example Terra Invicta has a neat version where you can make a traitor "councilor" automatically fail their dice rolls for he faction. But if do it too often, you make it too obvious they've been turned.

1 0 88 13K 16

View Details

Jonathan Fly 👾 @jonathanfly

4 months ago

@georgejrjrjr @_NathanCalvin I suppose there's some novelty in "Claude did all the research and made it happen, no experts needed" but tools that do this are widely available already.

0 0 1 19 0

View Details

Jonathan Fly 👾 @jonathanfly

4 months ago

@fofrAI I think it's actually both people, Anne Hathaway AND Nicholas Galitzine, a comparison: youtube.com/watch?v=2sT0mS…

0 0 0 490 1

View Details

Jonathan Fly 👾 @jonathanfly

4 months ago

@fofrAI @CauseItsOnTV Presumably there's an intermediate LLM prompt enhancement stage, just like Sora 2.0?

0 0 0 20 0

View Details

Jonathan Fly 👾 @jonathanfly

4 months ago

@janekm @fofrAI Great thread, the mundane examples are more convincing than most of the viral examples I've seen on Twitter.

0 0 0 27 0

View Details

Jonathan Fly 👾 @jonathanfly

4 months ago

@DanielleFong Check the history of posts. It's an ad.

0 0 1 311 0

View Details

Jonathan Fly 👾 @jonathanfly

4 months ago

I'm not sure I understand what you're trying to do. Cover an uploaded check and check the range of Covers you get back from Suno, as a way of trying to understand and see things in the uploaded track? I wouldn't recommend 0% weirdness btw, see how Suno's open source Bark at low weirdness can sound more weird than higher temps.

1 0 2 724 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

@Voxyz_AI @Alterverse_AI @Kling_ai third from the right morphs. though mainly I'd like to see improvements on scenes like this with the 8 characters interacting, when they are are independent it has the feeling of a videogame NPC idle movements instead of coherent at the full scene level.

1 0 1 16 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

@jasonyuan *Gossip Girl* as the positive vision of a future with humans contributing to a collective superintelligence... 🤨

1 0 2 545 1

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

@junmingong @j_stelzer Absolutely, now that the GitHub is out, I can answer all these questions. (I was just so curious as the details that it was frustrating the paper was so vague, since it wasn't out yet.)

0 0 1 20 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

@theo Hmn, gut feeling is the opposite. Having a pleasant but stupid model would make it hard for Google catch up. Usability/pleasantness can be iterated quickly, can do a lot in purse code with harness/prompts. Not everything, but much more than raw intelligence.

0 0 2 230 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

@myhandle Tokenization may be relevant, but some things work fine in a true base model (ie, song lyrics and rhyme schemes), so they must be a result of later training stages.

0 0 0 41 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

@levelsio I almost fell out of my chair when I ran this on my Intel 286 without a sound card, and this music and speech came out of the PC speaker.

0 0 3 283 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

@mweinbach In the thinking traces, to me it looked like it was using the equivalent of sub agents before. Same with GPT 5.2 Pro.

0 0 0 244 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

Even without the weights, if I could write custom sampling code and have access to raw outputs (tokens instead of decoded audio) that would be enough for me. But probably nefarious actors could use this so smuggle out the weights, not sure, I've never seen a commercial AI company offer that kind of middle ground access. For newbies, if they *just* expand beyond text only prompts and start using Inspire or other features, that makes a huge difference by itself. You can get a clear sense of what changes the funnel with some specific tests, like for example with "spoken word" there were times when it was difficult to keep Suno from bursting into song for more than 20 seconds into a song. Now you can both use sliders to increase style length, put other spoken word samples into the audio prompt, as well as generally prompting *anything* like literally covering "silence" will increase this "time to burst into song" metric.

1 0 1 49 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

I don't think you need that much access, the prompts (particular audio prompts) are strong enough to do something new in Suno. Over time you will the music get sucked into the "Suno Funnel" baseline, so you have to stay on guard and work in smaller chunks. But this was so much harder when Suno only had text prompts, you can overpower it now much easier than before.

1 0 1 53 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

@LaurensPlompen This one news.ycombinator.com/threads?id=jac…

0 0 0 14 0

View Details

Jonathan Fly 👾 @jonathanfly

5 months ago

Just caught myself replying to an LLM on Hacker News. I'm sure better bots already avoid these classic AI writing tics. Dread it, run from it, Dead Internet Theory arrives all the same.