That was painful! I discovered this last week:

I was wondering how they would handle the "Mama, just killed a man" section... and they managed pretty well.

Ok, it wasn't clear to me you were only talking about one piece of software. Thanks for the examples of evolution of digital sound (including in your addendum). Cool stuff!

People are flipping out over Suno and others not only because of ease of use, but it really is in a whole nevel level. Creating whole songs with vocals feels not like just a different color but a whole different horse. And yes, I appreciate that it a step in a long story of music technology, and yes I've read all your arguments on the page and agree with most everything, but I still find it shocking and amazing and it has caused me to reimagine what the future holds, for good or for bad. People aren't wrong or stupid for feeling this way after seeing and hearing what Suno and other current software can do. It is a sign of something amazing coming.

Speaking of digital instruments and waveforms, I've recently joined an orchestra and had my first concert this week. I'm mallet percussion. It occurred to me that this was something that digital music would never replace - 40 people working together on brass and wood instrument creating analog sounds in fantastic patterns. Nope, even when the tools and the AI are making excellent original songs in seconds, there will always be a place, a need and a want for human music.

Oh sure, and there is definitely the impact of shared culture is important. When I'm breaking out the guitar at parties, people would much rather hear songs they know and love than one of my own creations that no one knows, unless that is song is really good and resonates with them. But I've written songs that make me cry when I play them, and there aren't too many songs written by others that do that. Having a song that speaks directly to my experience and the way I've experienced experiences can be hugely impactful, and my own songs can do that me more than other songs.

So could an AI-song do this? Could it have the same impact, or would knowing that is generated by an emotion-less machine prevent me from connecting to it in the same way? That's a different question indeed – not a question of the content or composition but because of the connection to the artist behind the song. I don't know. It's a strange territory we're entering.

I saw "reasons" because I do not find your reasons especially relevant. I see you pointing to today's technology and saying (I think) that we will never to get to the point that AI can create as well as humans, and that doesn't fly with me.

When I say point-blank "You're looking at where AI is now and saying it will never get to human level" you repeat your claim that LLMs will not get us there. Not my point.

So let's be clear about what you are saying. Are you saying AI will never produce music as good as human-produced music?

Those videos are great. Thanks for those!

Yes, a big part of what makes 'Ghosts of Hubski' so good is that is personal and we relate to it. That a big part of what makes music good. Even some songs we love are because of lines in them that we relate to, even if the meaning we put to those words is nothing at all as the artist intended.

And that is what will also make AI music so good ... we will be able to produce songs to match our mood and experiences. Maybe no one else will like them, but I'm looking forward to the day I can get home and say 'Play a song to put me in a good vibe' and my computer will create a song to my liking with words that speak directly to me. It will be FANTASTIC.

You speak a lot about how you know what good music is, but don't forget that there is very individual taste. My big revelation long ago was that something is good art if I like it. And that makes me an expert on what good art is as experienced by me.

Yes, guaranteed by me. ME.

Sorry, but your arguments here and in the other thread simply display a lack of vision. You're looking at where AI is now and saying it will never get to human level because "reasons". You continue to cite your expertise and say that because you're an expert in a core related field that you know the future of how AI will affect that field.

I say you're most likely wrong. You'll say my vision amount to science fiction. But if you told the guy who invented the printing press that one day people would print lifelike color pictures with tiny drops of ink using a device that fits in your backpack and costs the same as a peasant gets paid shoveling shit for half a day, he'd say it's impossible. But of course it was not impossible. He was only saying that because he lacked vision. He may have been the most expertly expert on printing at the time, but he would have been wrong. (Disclaimer: I never met Mr. Gutenberg and have not discussed this with him, so I do not know if he was open to ideas of the future.)

Just because Sumo hasn't improved, does that mean that no other programs will improve? Just because the software right now is tying prompts to elements in a library does that mean that every single program from now to eternity will do the same thing? Of course not.

Yes, I know you're arguing that AI can only get us 80% of the way to what a human could do. I say it will get 100% there.

We could back and forth forever on this, but the thing is neither of us can prove the other wrong, so we're stuck with me saying you lack vision and you saying that because you have a long resumé you are more qualified to predict the future.

But what we can agree upon right now is that the people saying "We're cooked!" are overreacting. That was, in fact, my opening statement. We have different reasons for thinking that: I believe you are saying that the technology will never get us there because humans will always be more clever, creative and inspired, and I say that tribalism will decree that humans will value human creations over AI creations.

Artists are not doomed. BUT the AI-generated will continue to get better and better, guaranteed. To claim it will not be as good as human-generated music shows a lack of foresight.

mk's Ghosts of Hubski is an excellent example. A beautiful and touching song generated with 30 seconds of work using today's technology. And yes, I've listened to it several times. I'm not likely to listen it many times over a long period, but we are talking about something that would have seemed impossible ten years ago happening right now, with access to online tools anyone can use. In one year or five we will see gains upon gains... how can that NOT happen?

But human artists will not be doomed - how? I expect that live shows will become more popular. I foresee a rising anti-AI sentiment where it will become not cool to like AI-songs.

Expect to see progress and retaliation in waves. There will be human artists using AI secretly to create songs which they then play themselves. Arguments about what was AI-generated and what was not. AI videos of humans playing these songs which will be indistinguishable from reality. This will lead to live performances as the gold standard for musicians and fans.

It will happen. One does not to be an expert in music to see this happen. One only needs imagination and the ability to extrapolate.

I made the beep quieter - what do you think? Good sound level? Can adjust more +/- ...

Wow, that's similar. Slightly different rules, and I think Last Word has way better UI if I'm allowed to say so myself!

some days I really do. thanks steve for all your contributions! you've been a rock!

mk was the motivating force for much of this game - it was mk's idea and steve and I argued over rules. we all play tested the crap out of it. I like it better than wordle.

Thanks! mk had the initial idea after steve and mk and I were sharing wordle rounds and I got this one:

Wordle 1,208 4/6

⬜⬜⬜⬜🟩

🟩⬜⬜⬜🟩

🟩🟩⬜⬜🟩

🟩🟩🟩🟩🟩

WordleBot

Skill 85/99

Luck 57/99

No yellows today!

and mk got this one:

Wordle 1,208 6/6

⬛⬛⬛⬛⬛

⬛⬛⬛⬛⬛

🟩🟩⬛⬛🟩

🟩🟩⬛⬛🟩

🟩🟩⬛⬛🟩

🟩🟩🟩🟩🟩

no yellow here either!

We all play hard mode, and sometimes you get these long runs where you are just changing a letter or two. mk suggested making a version where you try to make as many words as you can without repeating letters in a position, and the next week I had some free time so I made a standalone app that was very rugged but it worked.

It was so much fun! We experimented with rules, and then it sat there for a while until mk prodded me again and suggested making an online version, which I had never done before. I've learned a ton!

Wow! Great score. Glad you like it! Still making a bunch of small changes...

I just started collecting data. The median for today is 7. Someone got 14!

Most days I get around 5-7, but I've gotten 12 before.

Image link is broken. Is it the same piece every time? Is it entirely over the calendar, or partly on the frame?

Thank you! But now I've been trying for 5 minutes to get a piece to float over the board and not settle into place and can't do it. I'm sssuming it's got to have its corners exactly centered in the middle of the squares. (If you're pranking me, well done! I'll be here all evening if I have to! Otherwise... challenge accepted! A puzzle within a puzzle...)

... and we're changing the arrows to unidirectional like you suggest! But keeping them at both the top and bottom. (And making them disappear after 3 seconds of not touching a piece). They can still be dragged, which is how I like to rotate them.

Thank you so much! We really appreciate you sharing! :-)

Thanks wasoxygen!

• Yes, we should change the arrows. I drag them in the direction I want, tapping them was a later addition because we found some people preferred that, but it can be confusing

• For consistency, I think we'll keep the flipping the rectangle option. I kind of like it!

• We're aware of vibration resetting. It will be fixed in an update going up today!

• We occasionally find an irreproducible error and it drives us insane!

• version 2 coming early next year will have social sharing that shows what frame you used without showing the solution ... it will be a little tricky to pull this off!

• version 2 will also have the ability to connect with friends and compete in weekly competitions, with a leaderboard and everything. Excited about this!

• we'll add notifications soon, too! Like, at 8 pm if you haven't solved for today it will send you a reminder.

.

.

Sven just sent a new build, so updates will be happening today or tomorrow! They usually take a day or two to get approved. Updates include:

• the solution bank will show the puzzles with the custom set you solved with!

• time machine when active will show "Yesterday" with "tap to reset" under it, to make it more obvious that you've got time machine going.

• some font and text size fixes

• saving light/dark and vibration settings

Nearly finished building an extra room on a small annex. Fun time working with a knowledgeable friend, we built it from the bedrock up. Scraped the earth, build foundation walls, leveled gravel, placed insulating sheets, poured concrete floor, framed it, translucent roof, windows, doors, siding largely done.

Still to do: finish siding, electrical, insulation, and paneling. It's going to be work room for our puzzle-making business.

Very satisfied.

Attach to a headband. Tell people it's an early Google Glass prototype.

When I was on chemo, the drugs would keep me awake for 3 days at a time. I lost the ability to make sense of numbers, alarming for me as a mathematician who can remember tons of numbers because each is like the face of an old friend. I sometimes lost the ability to make sense of time, I could not remember if something happened a day before, a week before, or months, and I could not sequence past events. steve made a short video about that strange experience.

After treatment, I slept terribly for a year and my thinking and memory were always foggy. Lack of sleep make a huge impact on my intelligence during that period (however I was extremely production and creative during that time as well. Weird.) I had resigned myself to my condition as my new normal, but my new new normal came around to be my old normal.

It is a kind moose. You can see him coming from far away, which can give great clarity. In any case, should you get covid and have your IQ drop a standard deviation, at least it would give the rest of us a chance to catch up!

Thank you so much! There is a free "explorers guide" coming out any day. 60 pages of cool stuff related to Abacabax. Just have to make the cover, but we got a big rush of business in our puzzle company and everything else is on hold at the moment.

Fantastic! Going to add color?