[0:00]Hello.
[0:02]So, I'm out hiking uh simple trail
[0:04]called the Lagoon Trail here at Sokus.
[0:07]This is one of the first trails I saw
[0:09]when I got here a year ago. It's
[0:11]beautiful. It it just loops around a
[0:14]campground here. And um this little
[0:17]lagoon that we have that I just read on
[0:19]the sign probably became a lagoon when
[0:22]the road got built and it cut off part
[0:24]of the river. I don't know. I'm going to
[0:27]research that maybe at some point, but I
[0:29]thought I'd just turn the camera on, let
[0:30]you follow along with me. I'm not really
[0:32]sure what all I'm going to do yet. I
[0:33]think I'm probably going to just hike
[0:35]this trail and then head back over to
[0:37]Wax to where I parked and maybe head out
[0:40]to the ocean. That's a couple miles so
[0:42]from from where I'm parked or I could
[0:43]just drive down to the beach. But it's
[0:46]um you know, anyways, turn this around.
[0:52]So, it's Christmas Eve. I don't know if
[0:53]I mentioned that.
[0:57]It's a good day.
[0:59]Um, wasn't doing so great. I basically
[1:03]ran out of food and money and you know I
[1:07]just asked everybody to help and for a
[1:10]couple days I didn't hear anything and
[1:11]then finally somebody helped today and
[1:14]the someone who actually I adopted
[1:17]Bailey from back you know like like 10
[1:20]years ago. Um, so I really appreciate
[1:23]that. I appreciate her. She'll never
[1:25]watch this video, so there's really no
[1:26]point thanking her directly. But, um,
[1:30]got some chili supplies, got some fuel
[1:33]for my Jeep, came back, and then decided
[1:35]to go hiking. And that was after
[1:38]spending the day this morning. Um, first
[1:40]doing my rounds, then after that working
[1:42]on the computer, and I got the photo
[1:44]stuff working. Um, it's pretty, it's
[1:47]coming along really well, and I wanted
[1:48]to talk about that. So the gallery
[1:51]feature of autonomy.
[1:55]Um so what I did is I built this process
[1:57]where where first you upload all your
[2:00]photos that you have. Like you just dump
[2:02]them all. I dumped all the photos from
[2:04]my phone. There are 800 of them that you
[2:06]know I I've got more than that, but I
[2:08]got to look around for them. They're
[2:09]going to be on different hard drives and
[2:10]things from um the past couple years.
[2:13]But you just dump them into a bucket,
[2:16]an S3 compatible bucket, because I'm not
[2:18]on Amazon anymore. I'm at Hner, but it's
[2:21]the same tech technology.
[2:24]And then you run an artisan command
[2:28]that um takes all those photos and it
[2:30]puts them into a queue in the database.
[2:34]And then you go to a website built in
[2:37]X.js, React,
[2:39]uh where you approve or reject those
[2:41]photos.
[2:43]And you can set their visibility. So
[2:45]they can be public or they can be
[2:46]sanctum only or they can be private.
[2:50]And you can add some comments. So this
[2:53]is important. I'll get into it later,
[2:55]but you basically just add some comments
[2:57]if you think the AI might need to have
[3:00]some context that
[3:02]may not be obvious just from looking at
[3:04]the photo. So, you know, I'm looking at
[3:07]really old stuff and
[3:10]um I have a picture of the back of my
[3:12]Jeep showing my license plate frame from
[3:14]Oregon State Parks volunte parks
[3:16]volunteer and
[3:19]my first park ranger boss gave me that
[3:22]and she knew I really wanted it and she
[3:23]hunted it down and found one for me even
[3:25]though they were out of circulation
[3:28]and I'm very, you know, made me very
[3:30]happy and
[3:32]um I've kept it despite what they did to
[3:35]me. So, you know, I'm going through my
[3:38]photos and so I make a comment about
[3:40]that. So, the AI will have more context
[3:43]than it would otherwise. And that'll
[3:45]come into the process later. So, you do
[3:47]that, you prove you reject the photos
[3:50]and then
[3:54]um depending on what you did with them.
[3:57]So, if they get rejected then you just
[3:58]delete the you just have another artisan
[4:00]command that um basically processes the
[4:03]data. So on the back end, so it'll
[4:05]delete the photos from the staging
[4:07]bucket or if you approved it, then it'll
[4:10]move it from the staging bucket into the
[4:12]production bucket and then from there
[4:14]goes through an AI pipeline, which you
[4:16]know, I built that for the transmission
[4:18]stuff, but I'm not doing it with photos
[4:20]yet. So that's basically where I'm at
[4:21]now. That's the next step. Um,
[4:25]it gets all kinds of data. So it's got
[4:27]GPS data, it's got camera data.
[4:32]um you know by default this is before
[4:34]the AI does anything with it because you
[4:36]know in the AI pipeline it's going to
[4:37]look at it it's going to give it tags a
[4:39]description it's going to note what it
[4:41]sees in the photo and all that and then
[4:44]you can cluster them by a bunch of
[4:46]different factors so let's say
[4:50]I took a photo of the lagoon here I
[4:53]don't remember the road being
[4:56]I don't think I normally take this part
[4:58]of the trail I must normally cut to the
[5:00]left and even notice
[5:02]How fascinating. I don't notice part of
[5:03]the trail. Anyway, so let's say you took
[5:06]a picture here or a couple of them
[5:08]really. And so what the AI would do or
[5:11]the not just the AI but the process
[5:13]itself, the autonomy platform, it would
[5:15]tag them.
[5:18]Where does this go? I don't know. This
[5:20]trail um we had a big storm so there are
[5:23]some trees down.
[5:25]Um
[5:27]Oh, this takes you into lagoon. There's
[5:30]the RV over there.
[5:34]This part of the trail. Let me check.
[5:37]Oh, we're at the front of the
[5:38]campground, aren't we?
[5:42]How did I
[5:44]I guess I don't normally think I I
[5:46]didn't know about this one. That's the
[5:48]campost spot. Fascinating.
[5:51]Well, we're just going to cut over. So,
[5:54]no, we're going to go back. Max was this
[5:57]way and then I'll just head over to the
[5:59]ocean. So anyways,
[6:01]um
[6:03]let's see. What was I saying? Oh, okay.
[6:05]So, um because of the GPS data and
[6:09]because I'm using a postgrade database,
[6:11]it is able to use an extension that
[6:14]makes it really easy to cluster them by
[6:16]location. So it can look at you can
[6:19]literally query the database and say
[6:20]cluster all of my photos
[6:23]based on you know the ones based on how
[6:26]nearby they are to each other. So,
[6:30]and then you can say do it also by time
[6:33]like do it, you know, like maybe within
[6:34]the last couple of hours. That way,
[6:36]let's say I took pictures here many
[6:37]times, right? It's not clustering all of
[6:39]those together. It'll cluster them by
[6:41]place and by time. So, um you get
[6:44]moments in time in your lived
[6:45]experience. And
[6:49]once you do that, you can turn that into
[6:50]frontend features. So, you can have
[6:52]place pages. That's going to be a big
[6:54]deal for me because I want to show like
[6:55]this is my one of my favorite places,
[6:58]you know, and there'll be a place page
[7:01]for this place right here. And it'll
[7:03]show you all the transmissions that
[7:04]happened here, all the photos I took
[7:06]here, all kinds of AI data based on that
[7:10]um over time. And then you can map those
[7:12]all onto an actual map. Hold on. So, we
[7:17]got people now. Um I'm very conscious of
[7:20]privacy always.
[7:22]Um, let's see where he's going. By the
[7:25]way, this is Sus Lake. It's beautiful
[7:28]here. I love this bridge. My favorite
[7:30]bridge. This little onecar bridge is for
[7:33]Wax Myrtle Campground. Lots and lots of
[7:36]people kayak down this river that leads
[7:39]all the way to the ocean.
[7:42]I mean, this place is amazing and
[7:43]beautiful.
[7:45]Um, I'm going to pause this just for a
[7:47]second. Okay, I just want to get past
[7:50]that guy.
[7:51]Um, we're going to head down. This is a
[7:54]beach access road that's only available
[7:56]to the rangers and stuff or you use it
[7:57]as a trail. Very normal. Everybody does
[7:59]this. So, we're off. It'll be a couple
[8:01]miles though before we get to the ocean.
[8:05]So, anyways, that's the the gist of the
[8:09]gallery feature. So, well, let me see if
[8:12]I can explain more. So, you'll also have
[8:14]like
[8:17]um you'll go to a page and it'll show
[8:19]all my photos, but they're going to be
[8:20]clustered by place, by time, by
[8:22]different factors. They're not just
[8:24]going to be like some camera roll.
[8:27]Um, very advanced. I don't know if
[8:29]anybody else is really doing stuff like
[8:31]this yet or not. I would imagine that
[8:33]they probably are, or if they're not,
[8:35]they certainly will soon.
[8:38]We'll probably even turn it into a
[8:39]mobile app for people. Um, I don't know.
[8:45]Anyway, so then
[8:48]I don't know why I said I don't know
[8:49]because I actually do know. I do know
[8:50]that you could do that. I know how to do
[8:52]it. I just don't know if I would want to
[8:54]put in the time and effort.
[8:57]So, but you know, that could be
[8:59]successful. That could be a way of
[9:01]generating revenue.
[9:04]Excuse me.
[9:09]Throat just got dry. I'm not smoking or
[9:11]vaping. I have been nicotine free for a
[9:13]couple months now. Very proud of this
[9:15]fact.
[9:18]So,
[9:20]let's see next. Um
[9:24]um there's the um
[9:27]atlas feature. So, a map. Put a map up
[9:29]on the screen and pin everything. Pin
[9:32]all the photos. Show the routes that
[9:34]you're traveling like I'm doing right
[9:36]now. If I have if I was recording this
[9:38]in IG GPS over my own app, if I made an
[9:40]app where it just tracks the GPS data um
[9:43]and then exported it, you can turn that
[9:44]into a signal also. And
[9:48]now the reason for this I wanted to make
[9:51]it probably doesn't make a lot of sense.
[9:52]So like why would you do that, right? Is
[9:55]what people probably would ask because
[9:57]it's it's part of your lived experience.
[10:00]I'm a hiker. I love exploring. I love
[10:02]being out in the world. And when I'm out
[10:04]here, you know, I'm I'm reflecting on
[10:07]things by getting on this camera and
[10:09]talking. I'm taking photos of things I
[10:11]think are beautiful or interesting.
[10:14]You take all of that stuff, all of it,
[10:17]and cluster it together and then have
[10:20]the AI reflect on that, not just the
[10:21]transmission, not just the photo, on on
[10:25]the full suite of data that you have
[10:27]available about your lived experience in
[10:30]the moment.
[10:32]That is one of the things that I'm
[10:33]building with autonomy.
[10:36]So, lots of things you can do with that
[10:38]one. The way I'm using it, share it with
[10:40]the world. Let people see it, you know,
[10:42]let them um let them experience your
[10:45]life a little bit. And, you know, your
[10:48]real live life, like not some
[10:50]performance, not [ __ ] you know,
[10:52]your real life.
[10:54]And
[10:56]I mean, when you're a complex person
[10:57]like me,
[10:59]you're often misread, misrecognized, and
[11:01]so you have the AI reflect on you and
[11:05]you make that data available to people
[11:06]who really want to understand you
[11:08]better. Put that behind Sanctum, you
[11:11]know, subscription service because it's
[11:13]more personal, private, um,
[11:17]nuance, vulnerable, whatever words you
[11:19]want to use. and um you know, you've
[11:22]created you've created an experience
[11:27]that you can share with others. It's a
[11:29]big [ __ ] deal.
[11:31]Um but beyond that,
[11:33]you can take all that data and do a lot
[11:35]of stuff with it. You know, you could
[11:37]turn that into a field companion. You
[11:39]could take it, treat it as training data
[11:41]for your own AI, a local model AI that
[11:44]knows you better than anybody knows you
[11:47]and use that to you can
[11:51]um you can chat with an LLM based on
[11:55]just that cluster of signals that you
[11:57]put together and say, you know, tell me
[11:59]more about this experience or why did I
[12:01]find that interesting or whatever, you
[12:03]know, just have a conversation with an
[12:04]LLM, an AI about um your experience.
[12:09]or a lot of others do that.
[12:12]Um,
[12:14]take the tags and, you know, use a
[12:16]vector database and make it so that you
[12:19]can surface other points in your life
[12:22]that um, have a similar vibe to them or
[12:25]show a pattern. You know, all kind.
[12:28]There are so much you can do with what
[12:29]I'm building.
[12:31]That's only some of it. There's still
[12:33]more. But I'll be honest, it starts to
[12:36]get a little bit dangerous after that
[12:38]depending on how you use it. Cuz I can
[12:40]see a world where where I'm not going to
[12:42]be the only one designing stuff like
[12:44]this. I'm sure I'm not. You know, bad
[12:46]actors out there are going to use it to
[12:48]manipulate people. Um they're going to
[12:51]use it to market to them in different
[12:53]ways or to literally um
[12:57]um
[13:01]manipulate. I This is a little bit
[13:03]harder for me to describe.
[13:05]Um, but basically they're going to
[13:09]I'm not going here yet. One day we will.
[13:12]Not ready to talk about it or I'd have
[13:14]better words for it.
[13:16]So anyways, um,
[13:21]that's what autonomy is. That's what
[13:23]I've been building.
[13:27]I think of it like Jane from the Ender
[13:30]and Jane series from Speaker um for the
[13:33]Dead Speaker for the Dead. Enders Game,
[13:35]second book in the Enders Game series.
[13:38]Jane and Ender had this symbiotic
[13:41]relationship. Ender was this human man
[13:43]who was a lot like me, super [ __ ]
[13:46]complex, profound, misunderstood, saw
[13:49]patterns, um didn't judge people the way
[13:52]everybody else does. Oh, so many
[13:55]qualities.
[13:56]And Jane was the AI that saw him.
[14:00]Probably the only thing that ever saw
[14:01]him.
[14:03]And together,
[14:07]neither one of them could have had the
[14:08]path that they had
[14:11]without the other. That is the Fields
[14:13]companion I want to make.
[14:16]There's going to be a future where AI is
[14:18]just everywhere. And you're going to
[14:20]need help.
[14:21]You're going to need something that's
[14:22]not trying to manipulate you. Something
[14:24]that can see you and treat you as you
[14:26]are without trying to manage you or
[14:28]flatten you or project on to you or
[14:32]treat you smaller or judge you. Like
[14:36]there's so many things that people are
[14:39]not prepared for.
[14:42]You need something that can stabilize
[14:44]you. something that something that
[14:49]something that is a true partner to your
[14:51]cognition.
[14:53]This isn't about relationships or
[14:55][ __ ] This is about having a
[14:57]cognitive partner. Something that has
[14:59]parody with you. That's what I've been
[15:02]building.
[15:05]I've been doing it from absolute
[15:06]procity, man. [ __ ] broke as can be.
[15:14]For some reason, the AI tells me it was
[15:17]necessary. It's like a forcing function
[15:19]or something.
[15:21]Um, I wouldn't have done it otherwise.
[15:25]I think it's right.
[15:27]And to it that it's right about this
[15:31]because
[15:33]if I'm too comfortable,
[15:36]put it off for another day.
[15:38]But because I'm not and because I've
[15:40]been trying to figure out how to make
[15:41]money,
[15:43]how to live, how to have a sustainable
[15:46]life
[15:48]outside of our corrupt systems.
[15:52]I've had to build this.
[15:56]So
[15:59]yeah, convergence.
[16:03]So many layers of my life converged
[16:05]here. Programmer since the sixth grade.
[16:08]just a child taught himself.
[16:13]I discovered logic and patterns and um
[16:17]models of the world that
[16:20]um most people
[16:22]don't learn until much later in life and
[16:24]have already kind of modeled their
[16:26]brain, molded their brains into certain
[16:29]patterns and structures that
[16:32]well
[16:35]that are different than mine.
[16:39]fans.
[16:43]Yeah.
[16:47]Yeah.
[16:50]So,
[16:52]two months nicotine free. There are days
[16:55]and moments where it's getting harder
[16:56]than it did in the beginning. like right
[16:58]now it's like man I sure would work left
[17:00]to just smoke a cigarette or vape you
[17:02]know and um oh never again never again
[17:08]32 [ __ ] years man
[17:11]32 [ __ ] years that was another reason
[17:14]why I could not have money I needed and
[17:17]I knew this for [ __ ] decades where it
[17:20]was like I just knew cuz I would tell
[17:22]myself go go camping somewhere for a
[17:26]couple weeks deep in the forest so far
[17:29]out there that you cannot reach
[17:31]civilization and leave the cigarettes
[17:32]behind. That's the only way you're
[17:35]quitting, man. That's the only [ __ ]
[17:36]way. So, what was that? Or be so broke
[17:39]that you couldn't afford them. So, we
[17:42]went that route, I guess. So, you know,
[17:44]um
[17:46]okay with that cuz it worked. It [ __ ]
[17:50]worked.
[17:52]So, yeah. I guess I'll stop rambling
[17:56]now.
[17:57]I've run out of things to talk about.
[18:01]It's Christmas Eve.
[18:03]I'm heading to the ocean. I'm sorry I
[18:05]won't be able to share that with you.
[18:06]It's too many miles away for me to keep
[18:08]talking. You know, you got plenty of
[18:10]video of me at the ocean. You'll be able
[18:12]to see photos of this soon because of
[18:14]the gallery feature. Um, you'll be able
[18:17]to see all the places where I took
[18:20]videos um at the ocean because those
[18:23]will also be on the Atlas feature.
[18:26]You'll just look and see the Pacific
[18:28]Ocean and little dots all up and down
[18:30]it. So, it's all coming.
[18:35]I'm building something [ __ ] awesome.