It looks like it is at the 1:2:19 mark in the video.
This looks insanely good. I have about 4k notes in markdown in a folder going back to 2003. There are background research for articles, presentations, subject research, and daily notes. Tailwind running against that could be of real interest.
ah wow looks great - love the way you can plug in your notes or trusted sources and mine them.
I am a dyslexic Scrivener fan, have loved Devonthink and am a constant dabbler in zettelkasten. To be able to interrogate these notes would be amazing. As someone who struggles with writing the LLMs offer an interesting strategy, the future sure is going to be interesting. Keep up the great and important work :)
Your mention of Howard Rheingold is a wonderful blast from the past. I had a chance to work with Howard and team back at AltaVista, (yes, that was a while ago).
Now, if we could also bring back HyperCard, we’d be all set.
Thank you for all the effort you put in to these updates. It's much appreciated.
I've had an increasingly dim view of Google of late, but I cannot deny the incredible utility of Gmail and Gsuite, and yes, Google Search. But Project Tailwind is exactly what I've been looking for as a tiny author who hasn't to date, indexed all my disparate writing. That "all" I have to do is to get my various articles into a Tailwind folder on Gdrive... and then I can run queries against it (and new stuff that I just dump into Gdrive, wow... that's powerful tool.
Still waiting on receiving Beta access to Project Tailwind, which has been renamed NotebookLM. In the meantime I've discovered Claude.ai by Anthropic. I'm eager to test out similar tools.
What's great about this is that it brings back to the fore the art / science of 'curation'. We have 'notaries' and 'apositilles' to give legal standing to documentation, verifying the the reliability of what those documents purport to be. AI built on scraping garbage will still produce garbage, only much quicker and amplify the junk at far greater rates. If tailwind stacks up and enables a verified curation from the outset, then arguably our starting positions for this what is a very interesting wide scale scraping plagiarism tool, should be more fit for purpose
First you get me hooked on DevonThink from your blog post. Then realize my daily / weekly notes are a sort of commonplace book, so I shape them a little more like that. But, I have over 70k papers, articles, etc. in DevonThink now and it still works rather well. I also have +4k notes in a folder (also pulled into DT). I have a large project I started in 2008 that I've been working on that were it a book as a short version would be 1,200 pages (working on non-book ways to get it out).
Throwing Tailwind at this to ask questions and get summaries and cross domain adjacent possible collisions would be great.
BTW the YouTube embedding seems to be a little funky so if you're having trouble getting to the Tailwind segment, it's at about 2:19 in the video.
It looks like it is at the 1:2:19 mark in the video.
This looks insanely good. I have about 4k notes in markdown in a folder going back to 2003. There are background research for articles, presentations, subject research, and daily notes. Tailwind running against that could be of real interest.
Am seeing the Project Tailwind segment begin at around 1:05:00, when viewing the video directly on YouTube (if that matters).
ah wow looks great - love the way you can plug in your notes or trusted sources and mine them.
I am a dyslexic Scrivener fan, have loved Devonthink and am a constant dabbler in zettelkasten. To be able to interrogate these notes would be amazing. As someone who struggles with writing the LLMs offer an interesting strategy, the future sure is going to be interesting. Keep up the great and important work :)
(Shame I'm not in the US) :(
Patiently awaiting more news on Tailwind….
Your mention of Howard Rheingold is a wonderful blast from the past. I had a chance to work with Howard and team back at AltaVista, (yes, that was a while ago).
Now, if we could also bring back HyperCard, we’d be all set.
Thank you for all the effort you put in to these updates. It's much appreciated.
I've had an increasingly dim view of Google of late, but I cannot deny the incredible utility of Gmail and Gsuite, and yes, Google Search. But Project Tailwind is exactly what I've been looking for as a tiny author who hasn't to date, indexed all my disparate writing. That "all" I have to do is to get my various articles into a Tailwind folder on Gdrive... and then I can run queries against it (and new stuff that I just dump into Gdrive, wow... that's powerful tool.
Thank you for your efforts Steven.
I am so in love with Project Tailwind. As an artist, and dyslexic, if I could add photographs, it would speed up the process for me.
Still waiting on receiving Beta access to Project Tailwind, which has been renamed NotebookLM. In the meantime I've discovered Claude.ai by Anthropic. I'm eager to test out similar tools.
What's great about this is that it brings back to the fore the art / science of 'curation'. We have 'notaries' and 'apositilles' to give legal standing to documentation, verifying the the reliability of what those documents purport to be. AI built on scraping garbage will still produce garbage, only much quicker and amplify the junk at far greater rates. If tailwind stacks up and enables a verified curation from the outset, then arguably our starting positions for this what is a very interesting wide scale scraping plagiarism tool, should be more fit for purpose
First you get me hooked on DevonThink from your blog post. Then realize my daily / weekly notes are a sort of commonplace book, so I shape them a little more like that. But, I have over 70k papers, articles, etc. in DevonThink now and it still works rather well. I also have +4k notes in a folder (also pulled into DT). I have a large project I started in 2008 that I've been working on that were it a book as a short version would be 1,200 pages (working on non-book ways to get it out).
Throwing Tailwind at this to ask questions and get summaries and cross domain adjacent possible collisions would be great.
Fascinating. Cannot wait to try out Tailwind.
That embedded clip isn’t working for me - do you the time the - presentation starts
When viewing the video directly on YouTube, try going to around 1:05:00.