FLOSS Weekly 758 Transcript

Please be advised this transcript is AI-generated and may not be word for word. Time codes refer to the approximate times in the ad-supported version of the show.

0:00:00 - Doc Searls
This is FLOSS Weekly, I'm Doc Searls, and this week, Dan Lynch and I have our second straight table for two conversation about what's going on now in open source, and we go deep on this one because we're right in the middle of this open AI crisis. Doesn't matter if it's over by the time you hear this, or Sam Waltman left and came back, but there are all kinds of open source implications for this, because that was an open oriented board. The board that fired the guy and got fired itself when he came back is a board that had an original imperative to keep things open because it's a nonprofit. It was all about that. Did it matter? And not not so much as the need that we have now to get open ethics and technology and the rest of it into the AI world. It's a whole new world. What are the new licenses we need here? It's brand new and it's gigantic and we need to be on it, get on top of it. So we go into that in this show and that's coming up next.

This is Floss Weekly, episode 758. Recorded Wednesday, november 22nd 2023. Raiders of the Lost Source. This episode of Floss Weekly is brought to you by Kolide. That's Kolide with a K. Kolide is a device trust solution for companies with Okta, and they ensure that if a device isn't trusted and secure, it can't log into your cloud apps. Visit to book an on-demand demo today. And by Bitwarden, the open source password manager, to help you stay safe online. Get started with a free Teams or Enterprise plan trial, or get started for free across all devices as an individual user at And by Fastmail, reclaim your privacy, boost productivity and make email yours with Fastmail. Try it now free for 30 days at Hello again, everyone everywhere.

Hi everybody. I am Doc Searls, this is Floss Weekly, and this week, both of us wearing hats one for a good reason, one for not Dan Lynch and I are for the second straight week having the Doc and Dan show and because we are guestless. There are different reasons each time, but here in the US it's Thanksgiving, we still have diseases going around, our guests got sick, so it has a good reason for not being in school. But here anyway, I am in the middle of Palm Desert, california, and there's very, very bright light out there, with the desert sun coming straight in and in the windows as curtains drawn, but they give a wonderful diffused light. The problem is I can't see the screen because the backlight is so strong. Anyway, so Dan. Dan is in Liverpool, where it is dark and probably rainy too. Is it drizzling outside, dan?

0:03:29 - Dan Lynch
It's pretty. It's very dark. It's too dark for me to see if it's raining or not. That's how dark it is. Yeah, it's good to be back there. It's good to be back. This is the second week of our Table for Two podcast Table for Two.

0:03:40 - Doc Searls
Yeah, and it's around Table for Two. Around Table for Two, that's true. As a coincidence would have it, the table on which my laptop is sitting is round and it is meant for two. It's a small one. Only two seats go for it, so it is consistent in that sense. But we have lots to talk about. That should be of enduring value and relevance and we're going to try and keep it to that. Okay, so I'm seeing in the back channel. I'm not heard on the live stream, but I'm going to pretend I am heard anyway and the live stream can go figure it out. Anyhow, a bunch of topics lined up. I'm going to start with one that Dan has already in the queue. He's the more homework than I have because I don't have Thanksgiving in the UK. There's actually two reasons for that.

0:04:37 - Dan Lynch
One is we don't have Thanksgiving, as you said. We're not thankful. And the second reason is because it's later in the day for me, so it's currently 5.30 in the evening here for me, so I have a whole day to prepare.

0:04:50 - Doc Searls
So I have a fair advantage really, and here in this time share condo space that we have. I am the barista. We actually bought, brought a cappuccino espresso machine here so I could apply my skills as a barista, Because when everything else fails I just go back to that. But I'm working on it and my wife is a very large family and lots of adult kids and other kids. I see some games over my right shoulder here and we had to kick all those people out. So anyway, go ahead, Dan.

0:05:33 - Dan Lynch
Dan. So I'm going to talk about what's really interesting about Red Hat Linux. It's a very popular enterprise business Linux distribution, but they recently this year changed their licensing model into a lot of caused a lot of upset with a lot of people because they basically made it so you can see the code but you can't do what you want with it anymore. So it's a kind of a show you the code and then look but don't touch is what people have described the peekaboo kind of thing. Yeah yeah, which which is obviously it goes against the principles of of open source, free and open source software. So but interestingly, alma Linux released 9.3 last week, which is basically a compilable equivalent of Red Hat Linux, enterprise Linux, but without all the branding and the and the other stuff kind of stripped out. So but the news this week is that a week behind them as our co-host Simon notes in our in our Slack back channel, they are a week behind Alma Linux, but Rocky Linux and also Oracle Unbreakable Linux have also now hit 9.3. Now that may not seem that interesting, but the interesting part of this is that there's a lot of business involved here, particularly Oracle. Because Oracle is such a huge enterprise company, they're selling their Unbreakable Linux, as they call it, which is basically a copy of Red Hat Enterprise Linux. They're selling that to customers along with support contracts and other things that Red Hat would like to be doing, and they're trying to eat in into their business in a way. But the interesting part of this is that Red Hat have not really done much to try and stop people from copying them in in, although they've made the license change, which we all well, I say we all I certainly don't approve of. But you know that's for me to say, that's purely my opinion. But they've. They've made this thing. But in the article I read it's from the register. In fact, the article I read about this, the release notes. They were talking about the business models of Oracle and other companies and how they're going to keep up with with all of this, and that's quite interesting.

So one of the one of the interesting things that comes out of the Oracle Unbreakable Linux release is they're using a newer kernel Then then you get with red hat Enterprise Linux. So red hat Enterprise Linux has a kernel which is based on 5.13 of the Linux mainline kernel and the new, the new, oracle released their own kernel with the distribution Sorry, 5.14. It's based on this, the. It's the roll kernel, the red hat Enterprise Linux kernel, but the unbreakable enterprise kernel, as they call it, and Oracle that is, that's based on a newer branch of the mainline kernel, which is 5.15 instead of 5.14.

So what people are speculating is maybe Oracle are trying to do a run around Round red hat by making their thing different, making their distribution Different and more valuable to customers by saying we've got a newer mainline kernel than you have. So it's just. An interesting thing is that we're going to see in the next few months and the next year's Probably it might take years, I don't know because enterprise doesn't move very quickly but we might see this business model Not go away. But it'd be interesting to see how they sustain that business model and how Oracle, for example, being such a huge enterprise focus company, how do they, how do they keep going with their, their business, which was essentially copying red hats product wholesale and rebranding it and then selling other stuff on top of it as well? So I don't know, doc, if you've got any thoughts on this, well do you a question.

0:09:41 - Doc Searls
I have it. It's actually a question for Dan, it's a question for our back channel, for the question for the audience is why did red hat bother to do that in the first place? I mean, here's a company that was founded on open source, believer in open source, we understood why they created fedora and like split redhead enterprise Linux and and Fedora apart. That was weird but it was like understandable because it had two different kind of broad purposes. But then why start closing down redhead enterprise Linux to the first place, especially if you had any foresight about it, you would see that hey, this is gonna get forked anyway and go off another direction, some of which are going to be competitive corporately. So why bother with that? And maybe there's an obvious answer, but I'm not sure what that is.

0:10:29 - Dan Lynch
No, I think you make a very good point there, because my my first thought when this happened was well, it's, somebody will fork it. I mean it's. We're not talking about small players here. We're not talking about people necessarily in a back room somewhere. We're talking about Oracle. Oracle, a huge. You don't think they've got the money and the resources and the time and the people to go.

Let's just put our own kernel developers and other people on to fork in this, you know, if they wanted to, because for them it's business. I mean it that they're a business, they're an enterprise. I mean so I don't know. I mean I can see why red hat would not want Oracle to be stealing their source, if you like, I do understand that.

But then again, if you believe in free and open source software and you've built your company on it for 25 years or whatever it's longer than that, isn't it? I mean, it's a long time. Anyway, if you've built your company on that for a long time and made a lot of money out of it from support contracts and always argued that you Don't need to sell the software because you can sell the support contracts and you can sell other things, to suddenly come and say you know, we've changed our mind on this and we need we feel that we need to be closing the you know the code now, or trying to hide the code away or change it in some way. Yeah, it feels a bit disingenuous to me.

0:11:44 - Doc Searls
I would say yes, especially with when you were saying that that Oracle are building this on a newer, the more current version of the kernel. You know, I mean it's one of the cool things about open source is that the kernel comes out, everybody can be in a race to use it the best way, and enterprises largely. It's a relationship cell, it's a corporate dependency cell. It's not, it's not a hey, we got the latest thing for your geeks to work on. It's we're building, we're keeping your company up Somehow and we're using this low quality Somehow and we're using this latest, the strongest of this. And you know, it's just, it's just do a better job.

You know, and I mean everybody, I mean in any country, everybody uses the same cash, they all use the same currency. That makes things simple. I suppose it's not a good enough comparison because anybody could create, fork the currency in some ways, but but it's, it's what's in common, that kind of matters and I was just, I was surprised by red hat when they did that in the first place, because it was Reputationally weird on top of everything else is like well, we've decided we're going to be closed in this way or we're going to be Closed fisted or closed vested on it. You know you're free to go and do your own thing with it if you want, but it just seemed. It just seemed strategically strange to me, and still does, given what's going on now.

0:13:21 - Dan Lynch
The other thing is the amount of as you said, the amount of reputational reputational damage that they've had from this. I mean, you've got now you've got a lot of people in the open source world of free and open source software community who used to look at red hat and go look there to shine an example of An open source company, a free and open source company, and how they make money from it and how they're incredible and they do this man. The piece that I read I meant I referenced it was on the register so people couldn't find the links, but it's basically one of the new pieces on there on the release notes. They were talking about the fact that red hat existed on an island, that the metaphor that they used was red hat existed on a very unique island when they were making money and they were doing well and they didn't care about the amount of reputational damage that they had from the Very unique island where they were making money and they were doing well and they didn't care about the little shoals of fish that were eating there. You know nibbling at their toes because it didn't make much difference. But now I don't know, is they're going to be a business model for the likes of Oracle to say we've got things that red hat enterprise and things haven't got now Like we've got this newer kernel, we've got this, you know, because one thing Oracle are good at and Oracle are An interesting company in a lot of ways.

I'll try not to give too much way and personal vitriol towards Oracle in this, but I mean, I'm not a huge fan of Oracle, but they do employ a lot of open source developers. They develop things like butterfs or betterfs if you want to call it that, and their version of the kernel. They offer custom versions of Linux kernel. They're offering versions of Linux kernel with butterfs included, which isn't in the red hat enterprise Linux kernel, and so they're finding ways to add value. And I don't know much about enterprise software and I don't know much about enterprise in general, but my opinion was, or my feeling was, that deals get done based on who plays golf with who doesn't. It Isn't that how that stuff works? I mean you, it's relational, it's a relationship business. So people work with the software sellers and tech support people that they know. So they're going to buy. If they work, if they're an Oracle company and they're used to getting supported by Oracle, they're going to keep going to Oracle as far as I can see.

0:15:27 - Doc Searls
Yeah, I mean it's. This is interesting, it's not even a tug of war. It happens at very different levels. You know, enterprises that employ a lot of smart geeks are going to make or lay that they respect and hold high positions and are responsible and involved, and all the rest of that will tend to hire on a big company, an Oracle, a red hat and IBM, which is also owns red hat and SAP. You know, based on some unqualitative judgments that are that are, that are considered. But then once they do that, a lot of responsibility for running the company goes over to these other companies. We see this all the time with. I mean, here's a, here's a simple example of a company gets hired to do a specific thing and then the company doing the hiring has kind of taken his hands off and not looked at it Every time. You see a cookie of sick cookie monster. But that's what it is, a cookie notice, you see is always from one trust or from one of the other companies that a company that a an enterprise will hire to handle. Oh, you can handle our consent stuff. You know, it happens with identity, it happens with security, it happens with a lot of a lot of areas where an external company is just going to be more expert at that than the company itself, than the enterprise itself. But with red hat and an Oracle, these are. These are big hires and I'm not sure they're made so much over golf, but they have to be. But they tend to be.

Informed is one of the things that got Linux into the enterprise in the first place is that a company gets into compliance with its own engineers. This happened around 2000 with IBM itself. You know IBM discovered that. You know they weren't running Windows print and file servers, they were running Samba, because there were lots of old Windows machines that work better on old Linux and got leveraged. And you know they're running Samba. And basically what I was told at the time is that IBM went into compliance with its own engineers. There was so many open source developers within the company that they just, they just went that way and it turned out to be strategically really really smart for them. But even internally and I've said this before, but Dan Fry was working there at the time told me it took IBM six years before they figured out that they couldn't tell their kernel developers what to do is really the other way around.

So, and I tend to think that there have to be, in some enterprises anyway, a degree of wisdom about this. But what kind of a judgment call is it to go? Is somebody to rip out red hot enterprise Linux and put in Oracle? Because of one, add on that, or because Oracle is one, something to the right of the decimal point ahead on on on the kernel is basing it on? I don't know, but again, I just I think it's.

I think one of the biggest issues that we're going to get to this later around open AI, because we have something to say about that is when is open best? And in our world, the floss world, the answer is all the time. You know that, on the whole, having having open source code and a collection of freedoms that are associated with that and are passed on to others is the better way to go and is going to give you more progress in the world. But there are, there are arguments against it, and one of those arguments wanted red hat at some point, and that is. It is disturbing because they were that, they were that that island, I mean and I've said this before too, but some people are new listeners on this topic.

But at Linux Journal, where I worked for 24 years, the very first editor in chief was Bob young who once he was writing about Linux, said wait a minute, I'm going to go start a company. And he started red hat and and the rest is history. But that history has been one of a growth and there are three stages in the history of a company there's, there's new and then there's hot and then there's big. And Red Hat got big. And when it got big and it got IBM, it went into a different stage. Bob Young was long gone. Other CEOs came and went and they made a strategic decision to do what they did with Red Hat, enterprise, linux and in retrospect I would say it wasn't the best decision, you know, and it cost them reputationally.

0:20:20 - Dan Lynch
Yeah, definitely, definitely. I think it's definitely something that's going to keep rumbling, but I thought it might be interesting for the listeners to have a look at the competitors and, as you've mentioned in the past, Al Mwellinix we've got people coming on in the future to talk to us about our Linux. So stay tuned to Floss Weekly to find out more about that.

0:20:41 - Doc Searls
And we'll be back with some fresh topics for our two-man table, for two round table, right after this. This episode of Floss Weekly is brought to you by Kolide. Kolide is a device-trusted solution for companies with Okta, and they ensure that if a device isn't trusted and secure, it can't log into your cloud apps. If you work in security or IT and your company has Okta, this message is for you.

Have you noticed that for the past few years, the majority of data breaches and hacks that you read about have something in common it's employees. Sometimes an employee's device gets hacked because of unpatched software. Sometimes an employee leaves sensitive data in an unsecured place, and it seems like every day a hacker breaks in using credentials they fish from an employee. The problem here isn't your end users. It's the solutions that are supposed to prevent those breaches. But it doesn't have to be this way. Imagine a world where only secure devices can access your cloud apps. In this world, fished credentials are useless to hackers and you can manage every OS, including Linux, all from a single dashboard. Best of all, you can get employees to fix their own device security issues without creating more work for your IT team. The good is, you don't have to imagine this world. You can just start using Kolide. Visit to book an on-demand demo today and see how it works for yourself. That's So, dan, what's next up here? I know I should know, but I have four windows open. Which one is it on?

0:22:34 - Dan Lynch
I think we need to talk about the big elephant in the room. I suppose the elephant in the room is the phrase, isn't it? Which is everything that's been going on with OpenAI in the last week or so. I've been trying to keep up with it, but I think you've got some interesting takes on this. Oh my gosh, I have a personal perspective. I would think.

0:22:55 - Doc Searls
Yeah, in another room here is my son, who is 27 years old. He's worked for a series of startups. He hated the startup life, went off to work in a restaurant. He's tending bar, having a great time now, as he puts it, every email he gets now is meaningful. He's not working in tech and he said he's back. He's back caring about it because of OpenAI, because of this drama and I know everybody is, I think, fairly familiar with this which is that OpenAI is run by a is a nonprofit set up by Sam Altman, elon Musk of all people who's since dropped out and some others with a nonprofit as a nonprofit, because they didn't want what they were working on to get into the hands of giant companies.

So it seems like the giant companies end up running everything and they didn't want that to happen. So and that may be over simplifying what they wanted to do, but they had this open imperative in the first place, which is why it's called OpenAI. But then, in order to, you know, buy the equipment they needed and rent the AWS whatever else they use in the storage space, it doesn't matter. They needed cash infusion. So they created a profit making subsidiary and that has grown faster than anything else possibly in the history of the world and if not in computing as well. I mean, this is 100 billion. I mean there's some crazy amount of sales that they're selling. I think sales are over a billion a year, and it was founded less than a year ago. As we're speaking, it's within a week of a year ago, and so the, for whatever reason the reason was never made clear the nonprofit board fired the subsidiary CEO, who was the most famous and high profile guy in all of AI, which is Sam Altman, and he had to leave his number two left with him. The entire company are close enough to it said we're going to go with the guy too If he's leaving the board of ports a, an interim CEO. She says that she's, she does. He wants to follow Sam out or bring him back. Then they hire a guy who is a twitch and nobody expected and he says some things that doesn't matter because. And then Microsoft comes in and says we'll hire Sam, and they hire Sam, and then, at the latest wrinkle in this is that Sam is now, as of last night, back at the company and the board is going to change and the original board is out, and all of the talk about this has been not just like what's what's happening next with it, sam, really do all that kind of stuff when.

Where I want to go with it is to the conflict between the imperatives of openness and the imperatives of a profit making company, a company that wants to create, you know, economic value for itself and and. But it's a little complicated here too, because it isn't just about the openness there there's these moral. I mean. One of the conflicts within the open source community, I think, is is it a moral, a morally bound community in which it's wanting to do good things, or is it simply about the value of openness and and what it does for the economy? And it has nothing to do with being especially good or bad? I think in this case both of those things are tied in. They, they. One of the reasons they created the profit making company.

As I understand it and I'm glad to be wrong here is that they wanted to be able to move ahead without having the drag of open sourcing this over and over and over again while they're doing it, and that there there always was an intent to open source as much as they could at a given point, and it was kind of an implicit, if not an explicit, promise that they would do that. Because they didn't do that, I think, is one reason why, in a limited way Facebook, which is a competitor, meta, which is a competitor, open source Lama or partially open source Lama and I know people there and they told me that they would have open sourced it all the way if they were not so competitive with other companies. It just didn't want other big companies taking their open source and using it, but they wanted the hackers of the world to take their, take Lama and use it. So all of these things are at play, but it's a, I think it's one of these moments when the economic and the technical and the moral imperatives behind the open source and free software movements are all in play. And I should add that somebody I know well and I think is familiar to almost everybody on here, but this is in a private meeting this person said, even though they were very, very in favor of open source and one of the great advocates of open source throughout its history, which dates, by the way, from 1998. At least the term, that term being used, free software goes back a decade and a half farther. But he said he deeply worried about what happens when Pandora's box is opened all the way and all the things.

Just imagine all the things that can ever be done with, with generative AI will be done, all of it, and the opportunities to do bad things are so far exceed the opportunities and the labor involved to do the good things, that bad things will surely happen and these are not like, hey, we're going to have the smartest thing in the world is going to be smarter than people and therefore we're all going to get killed, which I think is a logical leap. I've heard over and over again that everything follows. But if you can imagine, somebody gets into a position to make slaughter bots and make them personal and hacks into stuff that we could be known where we are in the phone company, whatever, and said slaughter bots, after all of us, there are all kinds of horrible things. You can imagine this person was imagining them and thinking, no, we need responsible companies or parties in charge of this thing to prevent that from happening. I'm not sure that's possible, but it's.

We're at one of those moments where and I am modestly called a thorough second law a second law says whatever it could be, anything that could be done with technology will be done. If it can be done, it will be done until we figure out what's wrong with it, and then we stopped doing that. We did it with nuclear power, we did it with all kinds of things. We're going to do it with AI too, but will it be too late? I don't know. What do you think, dan?

0:30:18 - Dan Lynch
Yeah, it could be. This is a huge topic. It's moving so quickly that I didn't even realize that Sam Altman had definitely gone back to the company.

0:30:27 - Doc Searls
Well, it was as of last night I got a deal with.

0:30:30 - Dan Lynch
Microsoft, but that didn't happen. Or does paperwork didn't get done. The really interesting thing about all of this is the you talked about. There are other competitors in this field as well, so it's not just open AI. I think they may. You know, clearly they're a long way ahead of some other competitors, but if you've tried things like I tried Bard, which is the Google AI, it's fine. That's pretty good. So are other things, so I'm sure the llama one's good as well.

0:31:04 - Doc Searls
Yeah, it's not the only game in town.

0:31:07 - Dan Lynch
So I think chat GPT is not the only game in town.

So it's a say that to restrict chat GPT to only be used for quote unquote good is an interesting one because you're not going to be restricting other things.

Google are going to do what they want with Bard and you know other things. And the other thing is it's wrong to think that the only people working on AI are people who work in Silicon Valley, because there are massive, massive countries. There's Russia, there's China, there's Korea, South Korea and North Korea and there's all these other places where they're going to work on this stuff and they're going to do whatever they want with it and we can't stop them. So I think, to use the argument that only you know, chat GPT has to be morally controlled. It's an interesting one Because for me personally, I have a lot of morals and I believe in certain things being done and I certainly don't want slaughterbots or whatevercold them coming around for any reason.

But I do think that you're right with your second law there, that whatever can be done will eventually be done at some point. And I think it's an interesting one because it's not just chat GPT. They're not the only game in town. It may be the biggest game in town, but they're not the only game in town. And the other thing about chat, GPT which I find really interesting, and open AI is that Microsoft already owns 49% of open AI of the company, of the for profit company. So I believe, although the structural as and pointed out in our back channel led, the structural kind of organizational arrangements of open AI are very, very complicated, but I do know that Microsoft have got heavily, heavily invested in it and things like Bing and also co pilot which we're going to talk about in a bit, because I've got another story to tell people about an uninsulated one and co pilot, which is used by GitHub, which is also owned by Microsoft, and other things. They're all basically powered by chat GPT based tools that they're getting from it.

0:33:07 - Doc Searls
For complexityai is really good and it's it's mostly chat. Gpt is 3.5. But it's very, very handy for all kinds of things. I probably use that more than any other AI. I should answer, as you said, that it isn't just not only is it not just open AI plus Microsoft, which is now essentially one entity, the irony there being that they wanted to avoid being falling into the hands of a giant company and a giant and with what this board did was basically hand open AI over to Microsoft. It's a follow the money thing. Right, you know where does the money come from? It came from Microsoft. Microsoft is a partnership. Microsoft runs this thing, so to speak. So, but I want to point out that I mean Google's.

I first found out about AI A deep, interesting things happening with AI from friends in the UK. Mustafa Suleyman is not a friend, but I've talked with him and he's one of the founders. Ben Laurie, who's a great hacker and I've known for many, many years, worked for them. They think he still does and they got acquired by Google deep mind and they're involved with, you know, neural learning machines and you know I mean they had a neural network to begin with and I assume that's involved in Bard. It's a I mean. This is going on all over the place and that's another reason why it's it's not like there's a cat out of the bag. It's like many species have jumped out and they're all over the place and there's more than one bag and it's it's more than one bag.

I mean it's, it's part of, part of. It is just the implication of, of the technology. I mean I I own almost no stock at anything, but I'm not even, I mean is handled by other parties, actually like my wife and other people. But I mean I wish I'd known how important Nvidia would be, you know, I mean, but the fact that Nvidia made the, the silicon and the and the products they made is one of the things that allowed open all kinds of AI to explode, and so the presence of those in the world really matter. Now AMD is getting is jumping in on it. There are, you know, our makers that are jumping in on it. They all want to come up with ways that that AI usage could be accelerated.

And the big question for me is when do I get mine here? Not where I'm, I'm inside of a giant company, but I've got the models and I've got my own data. I want to know, I want to look at my own health and stuff like that. But there is going to be this kind of weird symbiosis.

One of the things I heard on another podcast this morning as I was riding around Palm Desert is that in medicine already it's been figured out. I mean there is somebody's come up with something where an AI and machine learning system, with you know, we'll be able to look for the markers for pancreatic cancer far better than any diagnostician can, and the implications for medicine are absolutely colossal. The implications for countless things are colossal and and this is where we have to go to what can be done will be done, because it is and it's going to happen fast. And and I think one of the one of the things that's surprisingly absent for me is Apple. What's Apple doing with this? Nothing, as far as I know. I mean, I've had AI for a long time in various kinds. They've had facial recognition and stuff like that in their, in their, in their photo products and so forth, and I imagine they're not idle on this and they probably could buy some AI companies with spare change at this point, although probably the price for all of them has gone through the roof.

0:37:19 - Dan Lynch
Yeah, I think my my issue with the way the story was reported in. I say issue it's not, you know, it's not like I'm up at night going this makes me mad. But my kind of slight issue with the way the story was reported and the way people have talked about it is it goes back to the point I was making about they're not being the only game in town. There are other people working on AI outside of of that company who may come up with something better and quicker than than open AI. So we don't, we, we don't know what's going to happen.

0:37:53 - Doc Searls
And we're useful. I mean, you know, I go ahead.

0:37:58 - Dan Lynch
Sorry, sorry, doc. No, I was just going to say that we need to. We need to give people another message about from our aunts telling us we need. We need to give people another message, but after that people should should definitely come back, because I'm going to talk about an open source company that's doing AI stuff that relates to this. So, doc's going to give you some message, okay.

0:38:20 - Doc Searls
Thanks, dan. So this episode of Floss Weekly is brought to you by Bitwarden, the only open source cross platform password manager you can trust. Security Now's Steve Gibson has even switched over. With Bitwarden, all of the data in your vault is end to end encrypted, not just your passwords. Bitwarden protects you by creating unique user names and adding strong, randomly generated passwords for each account, or using any of their six integrated email alias services. You can log into Bitwarden and decrypt your vault after using SSO on a registered, trusted device. No master password is needed. On top of being public to the world, bitwarden has professional third party audits performed yearly and published on its website. You can even view all of Bitwarden's code on GitHub.

Share private data securely with fully customizable plans for personal or business needs. Start with a personal plan for free forever and get unlimited devices, unlimited passwords and pass key management. Or enjoy premium features by upgrading to a premium account for less than $1 a month. And for only $3.33 a month, their family plan will give you up to six accounts with premium features and unlimited sharing and collections. Their enterprise plan is just $6 per user per month and now they have a Teams starter plan, which is a flash $20 per month for 10 users, which provides a great value for smaller teams with all the premium features, or pay $4 a month per user with teams of 11 or more.

Join Bitwarden at the fourth annual open source security summit, where businesses, open source enthusiasts, developers and security thought leaders come together to explore how open source software leads to stronger cybersecurity. Headliners this year include Brian Krebs from, white Hat Hacker and CEO of Social Proof Security, rachel Tobak, and generative AI expert and advisor, zach Tass. Join the leaders in open source security and register now for the open source security summit on December 7th at Here at TWiT. We're fans of password managers. Get started with Bitwarden's free trial of a Teams or enterprise plan, or get started for free across all devices as an individual user at bitwardencom/twit. That's Okay. So, dan, go ahead with what you had queued up there.

0:41:12 - Dan Lynch
Yeah, of course, sure. So we're talking about AI. We always seem to be talking about AI, but it's the thing that everybody needs to talk about right now. So I want to talk about something called TabiML. That's TABBYML, which is a machine learning obviously tool, but it's an open source challenger to GitHub's co-pilot. So co-pilot is the tool in GitHub that helps you to generate code. It will go and grab things mainly from other GitHub repositories, so that's a strange one, so you kind of copy in other people's code directly from there. But anyway, you could be. It suggests things. Anyway, to be fair, I haven't used it. My friends have used it and told me it's very good. But the story here is that TabiML, which has been created by two former Google employees, is a startup which is just generated, or just secured $3.2 million in funding to work on their open source code generation tools. In contrast to GitHub's co-pilot, a self-hosted coding assistant like TabiML has the advantage of being highly customizable.

Suggesting the startups founder Meng Zhang, we believe that in a future where all companies will have some sort of customization demand in software development tools. So it's just an interesting one, because they go on to talk about the fact that if you can embed the AI. It's kind of like we always talk about we'd like to have our own AI so that we can embed and examine our own data and give us our own suggestions based on that. If you're in a company where your company rules, as most companies are, if a developer writes a line of code, it's the property of the company. It's not the property of the developer. So therefore, if you've got your AI embedded inside the company, it could immediately quote and suggest things that other employees have written, rather than, say, co-pilot, which can't do that. Co-pilot doesn't have reach inside your company because it's not embedded inside your company. There's all kinds of legal stuff there, I would imagine yeah, copyright things and licensing things.

So I just the story I thought was interesting was it's good to see that other tools are being created, and you talked about complexity, ai and there are other ones as well. There's the hugging face, which you've mentioned as well, dark as it is, but I think open source AI tools are definitely going to be interesting, and I got the story from Yahoo News, by the way, so I should mention that from Yahoo News was where I picked up the story of this one. It's a little bit old. It's about three, four weeks old the story, but I figured, given that we were talking about AI, it was relevant. So I'm going to keep an eye on Tabby ML and see how it comes out. I also like the fact that they've kind of done a clever thing with the cats. So GitHub has OctoCat, which is their kind of logo, which is a cat and an octopus half hybrid. So they've called it Tabby because it's a cat. So there we go. I thought that was quite cool.

0:44:11 - Doc Searls
It's a cat with socks, with its socks on it. It's funny, yeah. And since people hugging face, there's another piece of news From way back in November 14th, which is like a week ago, that Dell infrastructure, of course, is like big Dell. This is enterprise type. Dell embraces hugging face and they have a deal on open source generative AI models. So open source is a big deal there.

And I should add and maybe we're getting ahead of ourselves here but one of our other co-hosts, simon Fiffes, who is listening in hi, simon, I wonder. Well, you could make this table round if you came in, or at least triangular. But he says it's unclear what open source means in the context of AI and most usage seems questionable. Only Eluthor seems to be trying it. Osi, which he's deeply involved with, is doing a deep dive in the topic and have added a solid definition which, if you go to opensourceorg or to blog that open sourceorg and you'll find it there, and they've come up with at least the first draft of a definition. They take their definitions very seriously. They've got a bunch of them and they're really important to keeping the conversation around this sane. There has to be. This isn't governance in the same of you get punished if things are right or wrong, but rather governance is what we agree on is right action of one kind or another. So, anyway, take a look at opensourceorg and dig down in there and see what's, because we need an open source definition of what happens here.

We've barely begun to scratch it, and so much of what's happening in the world is not just open source. It's just who's using it and how's using it and all the rest of it. I'm going through a similar thing right now with and this has to do with the status quo. I think this is important. Once a status quo gets established, whether it's good or bad, you have to deal with that status quo, and so I'm involved in a standards effort right now where we're trying to both respect and ignore what's good and bad about the GDPR, the CCPA in California, the DMA coming along in Europe, lots of law that's coming along, that, and regulatory frameworks that may be well intended or may not be, but if we respect them too much, we may move away from the other things we value. It's all got pretty complicated, but the OSI is really good at this stuff and has been involved in defining open source for many years.

0:47:07 - Dan Lynch
Definitely an interesting subject I actually have. On the subject of monopolies and so on, I have a question from my master Don account that someone asked me earlier. I asked for potential topics for the show and I had a reply from Tom, who doesn't give a surname, so thank you, tom, for that. He asked about monopolies. He says are monopolies that are open source thinking of FFMPEG here as the only media encoder used good or bad for the ecosystem? Is it open source so great? But is a monopoly or a monoculture still bad? So I don't know. I think it might be something we could quickly discuss.

0:47:51 - Doc Searls
Yeah, it's funny. A couple of days ago my son's girlfriend and I went to Joshua Tree National Park here in the US, which is where it's this valley of boulders that have these really unusually shaped trees that look like Dr Zeus Trithem. But the interesting thing to me is that there are a bunch of boulder piles that look so much like the XKCD cartoon, where the entire internet depends on one thing, maintained by one guy in Nebraska or somewhere, and with single sources of some conditional code that we're using become over important. That's part of it. I think that's part of where you're going with this. And there was one particular pile of rocks that had one holding it up.

And later, when I looked to see, I'm very curious about geology and it turned out these particular rocks are over 1.3 billion years old, but they're under the earth for most of those 1.3 billion years. Now they're out being used by nature and are crumbling and we're constantly replacing everything. I mean everything has to be kept up to date. I don't know, I'm kind of going all over the map with that, but I'm not going to go ahead, sorry.

0:49:15 - Dan Lynch
No, no, I was just going to say. It's interesting that I think putting all your eggs in one basket, as they say, is never a good idea. So I think, whether it be an open source basket or a closed source, proprietary basket, I'd prefer it to be an open source basket Still not a great idea, because you know, I mean the example he gives of FFM peg, which is for people who don't know, maybe we're listening. It's an. It's a media encoding tool, does video and audio and so on, so it does all that kind of stuff and it's basically the only one that everybody uses and embeds. So I think you need options. I suppose is what I'm saying. You need to have options.

0:49:55 - Doc Searls
Well, we're getting down close the end of this thing, and I have to get in one more message, so I'm going to go for that and then we can. We can sum up and wrap. So this episode of Floss Weekly is brought to you by FastMail. Make email work for you with FastMail. Customize your workflow with colors, custom swipes, night mode and more. FastMail now has quick settings. From the quick settings menu, you can easily choose a new theme, switch between light mode and dark mode and change your text size without leaving the FastMail screen you're looking at. Quick settings will also offer options related to the FastMail screen you're viewing. You can generate a new masked email address, show or hide your reading pane, switch between folders and labels and more. Choose to autosave contacts or choose to show public images of senders from external services like Gravatar, set default reminders for events, change how invitations are handled or turn notifications for calendar alerts on and off. Now add or buy a domain through FastMail and they will set up all the records for you. So it works immediately. FastMail gives you the ability to send and receive emails from your own domain and manage multiple email addresses in one space, which helps keep you organized and protects your personal data.

For over 20 years, FastMail has been a leader in email privacy. The FastMail team believes in working for customers as people to be cared for, not products to be exploited. Advertisers are left out, putting you at the center. You pay for free email with your privacy. At FastMail, your data stays yours, with better productivity features for as little as $3 a month. FastMail has better spam filters and absolutely no ads. And privacy isn't all you get with FastMail Superior productivity tools with scheduled sends, snooze folders, labels, search bar, etc. Plus, keep track of all the important details in your life easily with FastMail's powerful sidebar. Works with password managers like Bitwarden and OnePassword to make it easy for you to create unique passwords for every account and safely store them on your device. It is great on your desktop and mobile as well, especially when you download the FastMail app. To get the most out of your email, it's easy to download your old data and import it into your FastMail inbox.

FastMail is moving email forward with new internet standards and open source innovations that power many email services other than their own. Don't get left behind by substandard email providers. Reclaim your privacy and boost productivity with FastMail. Try it now free for 30 days at So, dan, it seems to be. There's a theme here and it's where does open source stand in? A world is changing almost way too fast. It clearly faster than it's changed in any other time I can recall, and that even includes when the internet showed up, I think open source is still at the center of everything.

0:53:33 - Dan Lynch
Certainly, I'm very biased in this, but when you look at so we've talked a lot. I was going to say we've talked a little bit about AI. We've talked a lot about AI today, but if you look at, you mentioned medical uses for AI. So examining diseases, you know, diagnosing things, looking at test results, all that sort of stuff. That's the science, and in science, peer review is everything, and peer review is all about being able to see the process, be able to recreate the results, be able to recreate the test that people have done.

That, to me, is all open source. You know, in a way, that's like an open source way of doing it, and I think that, although the world is accelerating quickly, I think the sharing of knowledge is still the key to human advancement. That sounds very grand. The key to human advancement is that it's a sharing of knowledge, isn't it? So I think that's still relevant. Open source and free software and all that sort of stuff is still very relevant because these things are going to be shared and whether you want them to be or not. So why not get on board and use the proper licenses and to try and share the software and the codes and the other stuff, codes, the code and the other stuff and then you can also you know you can benefit a lot from that. We've seen that in a lot of ways and we're very much here to promote the idea of freeing open source software. I think anyway at least I know I am and you know I think it's very important.

0:54:59 - Doc Searls
So I have a. This just came to me as we were talking, but I think it's important. I was thinking back on how the free software movement really and where this showed up in the first place with Richard Stahlman back in the like around 1984, at that time, I mean, he wanted to take UNIX and have an open UNIX, have a free and open UNIX that didn't have to belong to IBM, and then, for that matter, with BSD a little while later, that's what Bill Joy wanted to do as well, but he created free BSD as a UNIX variant, and the idea in both those cases and in Linux when Linux came along, was, as a guy, there was a person, an individual said here's a big corporate thing that's closed and is going to take forever to develop and is going to be locked inside some company and we're going to have to, like, beg them for something, and we can always think of better ways to make this thing work. We're going to do it on our own. And there was an I and then there was a we, and I mean what's the first thing Linux did after he created his first whack at a POSIX compliant UNIX like operating system was share it on the internet right and other people weighed in and then they made the history we've had since then and it's happened in so many other ways.

I think with I'm thinking like right now, with, with AI, if anything we're doing in our life, like right now, we're at a I'm at this. It's kind of a resort. It's not really. It's a whole bunch of time, shares, time share, condos, but it has a central administration and lots of events and the horseshoes and basketball and pickleball and and yoga in the morning and stretching and other stuff like that. But there's also stuff going on in the in in the region. There are plays and theater going on, there's a gondola you could take up to the top of a mountain. There are all these options that we have that are. Some of them are recreational, some of them are shopping. The Googles of the world will want to say we're going to know everything we can about this and we're going to tell you what we can, but it's not consolidated, it's not. I mean they're. You know, they sent us as good as their traffic and ways, traffic in their own ways, but they're different. Both of them sent us into a horrible traffic jam that we, at our human level, we could have avoided. In Los Angeles several days ago when we were driving here, because it was a road that somebody set a fire in but there was a big fire on the burn, an extraordinary, you know like seven lanes each direction, very busy road, and we were tied up late for 45 minutes to an hour just on city streets where they were totally gridlocked.

I think that if one has one's own AI where you can choose from all kinds of sources, you don't have to, you don't need the dependency I mean. That's basically. I mean I mean the earliest memories I have of working with Linux was check your version, dependencies right, but dependencies are everything right. It's a collection of dependencies. We don't want dependencies on giant companies alone. I know Amazon can get us stuff to me faster. I think there are. If I need a stapler, or if I need a, you know, a shovel, some other thing, I need a grill. I'm looking out here. There's inventory all around. It's not only at Amazon.

I want to be able to get at that and I'm an individual human being and I think other individual human beings working with centralized AIs, learning models of our own, are going to be able to do better innovations, better optimization of opportunities, of possibilities, of how we work together. Whatever it is, traffic is working together. Right, traffic has a mind of its own. It's like a blood stream. There are independent minds in there, but we're all operating independently as well, and that's one of the weird and interesting ways that humans are interdependent on each other, and I think having openness as a value and open code as a technology is essential for making the most of that and also of human beings' ability to help each other and to be you know, to be constructive with each other and not just opportunistic and try to beat the other guy. So anyway, that's sort of my sermon on that that I just made up on the spot.

0:59:57 - Dan Lynch
I completely agree. We need collaboration. I mean, you know people have done great things on their own, but it's no good if you don't share it, is it? It's no good if you don't tell someone. It's no good if you don't. You know someone else can't build on that knowledge. If we didn't have the commons and we didn't have you know all these things of like libraries full of books we can learn from you know all these types of things, where would we be? You need to share knowledge. Knowledge is dead if it's not shared.

1:00:20 - Doc Searls
in my opinion, yeah, and it's interesting I mean right now, though I'm not there in geographically. At the moment my wife and I are working with Indiana University on the commons at the Ostrom Workshop, which studies the commons. Eleanor Ostrom won a Nobel Prize for her work on the commons. She was familiar with open source and open source development and that's been a model for the way commons work for a while now. And we're still discovering what that means. And I think at the center of it is how could you know, how could we be at our most human while we're busy running the machines that help us in ways that we will never be that smart, we'll never have the memory a machine has. A human grace is that we forget most of what we're saying or heard within seven seconds. And yet we had. We. We pass a long meaning to each other with language and with mathematics and other things and with code. But the question for me is you know, how can we be most human in the midst of this? And we've got this explosive new science, not that new but capable in ways it never was before, and I really feel like it's 1973 or 1974, when computing was hot. Everybody knew about it. You know, just a few years earlier we had a Hal 9000 on 2001. And which is super smart, but the story was about how the human, you know, was able to get past what was wrong with it. But the we actually thought as a collective that the big company is going to do it all for us. Ibm was going to do it for us, but in the meantime, apple, atari, commodore, osborne, sinclair, all of those were happening over here. Nobody took it seriously in big computing. And yet that's what? As soon as, as soon as people got PCs inside of companies, companies worked much better and they totally stripped the gears of all the mainframes that were in there. And there's something like that going on now.

I mean, I watched the PC revolution happen. I watched the internet revolution happen, really exploding in the mid 90s, which ended up with the cooperation of every network in the world. They had to cooperate, you know, and all this cooperation happened, and now we have the internet, which is a miracle beyond the scope of loaves and fish. That eliminated, eliminated the distance between everybody. And then we had the cell phone revolution and we just took all the computing we made and we have an extension of ourselves. That are these things that we carry with us, but they're still kind of run by the big companies and Apple and Google pretty much run the run the mobile, the mobile world, and that's not. That's not a good thing.

And I think we're going to see something happening like that with with AI, where it's going to be meta, and Microsoft and Amazon and a few others, and you know, eventually Apple comes along too, but we'll do better with it on our on our own and we'll come up with innovations on our own that are based on open code that we share with each other and do cool things with. And, by the way, we haven't talked about this AI is incredibly fun. I've had, you know. I mean this is something that's lost in this. I mean it's, it's really handy. And the programmers I know oh, my God, you say great, I do that much Python, it'll work out. The rest.

1:04:10 - Dan Lynch
That's pretty cool. It is amazing. Have you done bash scripting with?

1:04:13 - Doc Searls

1:04:13 - Dan Lynch
No, I haven't, I haven't, I'm bascripting with, with, with AI helpers, is very, very interesting. A friend of mine was telling me he works for a logistics company they deliver packages, you know, and he was he's works in the IT department. But he was telling me that he just asked this was like a year ago or or a little bit less than a year ago, when chat GBT first came out, he asked it to to give him a bash script with which would do, I think, about 10 different things, and it just spat the thing out and he was like and it worked and it was perfect. Well, say perfect, it worked anyway for that length of script.

So, yeah, it's interesting times anyway, but I think I think we yeah, we've we've got interesting times to come in future with all of this stuff and the power. At the moment you need the a lot of computing power to do this and sadly we don't carry that round in our pockets. But then again, you know, 10 years ago I didn't carry around half the computing power I've got in my pocket right now. So who's to say when that will change? It will happen, it will change as things go.

1:05:15 - Doc Searls
Yeah, I assume we're. We're down at the end of the show and and I think we're ready to wrap it up. Yeah, I think we're ready to wrap it up. So thanks everybody for for hanging with us on this thing, and I do want to give a quick plug before we wrap it up oh yeah, please. So I was going to yeah no problem.

1:05:37 - Dan Lynch
So I mentioned it on a super secret pirate show that some people got to hear last week. I released some new music recently.

1:05:45 - Doc Searls
Right, this is great and I listened to it as good. Oh, you listen to it, oh, cool. Yes, of course.

1:05:50 - Dan Lynch
Yeah, so you can listen to it. You can find it at danlynchorg slash music. That's the easiest place to go. It's a new EP, four songs on there. They're all very different, so it's eclectic, I warn you. There's electronic stuff and there's other stuff on there, and you can also get some of the older music on there as well if you're interested. And it's all free as in beer, and free as in freedom as well, because it's released in the creative commons.

So if you go to band camp, you can. It will ask you to. When you go to buy the track, as it says, it will ask you what you want to pay for it. You can put in zero. That's a valid option. I don't mind that. If you want to give me something for it, you can do. That's also cool. But you really don't have to. You can. You can pay zero, but you can also get it under a creative commons license and remix it. And let me. The only thing I want to know is what have you done with the remix? You know, I want to hear it, I want to play around with it myself, maybe in the open source ethos, and that's yeah, that's cool.

1:06:48 - Doc Searls
Yeah, and I have to. I have to say there are our guests next week. Surprise, I have our guests hoping he doesn't get sick and and we don't run into the same problems again. So we don't have the Dicadan show whatever you're going to call it for the third week in a row is Evan Predromo. So he's the co-editor of Activity Pub, the W3 standard. Let's see if you're on Mastodon, you're using that. He was big with Identityca years ago. Identity and Statusnet did work with Fuzzy AI, artificial intelligence stuff. I think of him. I think of him as a Canadian, even though he actually is from Cincinnati. He's a fascinating guy.

1:07:39 - Dan Lynch
I've interviewed him myself in the past and he will be really interesting. So people definitely come back and tune in for that next time.

1:07:46 - Doc Searls
Yeah, he's great. He used to be. It probably will be again a fixture at the Internet Identity Workshop which I helped put on twice a year. Very involved in making things happen, he's always he's one of those people who's always pushing things forward. It'd be really great to have on the show. So and on that one I think Jonathan Bennett's going to co-host that one. At least he's lined up, unless he gets sick or something happens in Oklahoma. But in the meantime it's been a great show. Thanks a lot, dan.

1:08:20 - Dan Lynch
No problem. Thanks very much Doc.

1:08:24 - Leo Laporte
Listeners of this program get an ad-free version if their members of Club TWiT. $7 a month gives you ad-free versions of all of our shows, plus membership in the Club TWiT Discord, a great clubhouse for TWiT listeners. And finally the TWiT Plus feed with shows like Stacey's Book Club, the Untitled Linux Show and more. Go to and thanks for your support.

All Transcripts posts