00:51:58 · Luca Ferrari
[00:51:58] S0: Yeah. Alt Spooner as an ALT alternative or alter ego is just one of those 50 plus tools, something that we we developed this year, and it's it's it's been adopted very enthusiastically across the company. Basically, build this... It's really like an agent that lives in Slack, at least the interface with Spoons members of the core team. We'll be looking to deploy this across all of our team members, but you interact with it on Slack as if it were any other colleague, and you have one old Spooner, you can name it and give it a profile picture and all that. And automatically that agent will have exactly the same level of access you do across all of our platform. So if you have access to certain code base, it does too. If you have access to a certain customer support tool with a certain level of permissions, it does too. And I can basically task this agent to do stuff for you and it could in principle do pretty much anything, like anything you can do in principle, can do, obviously AI is not perfect, so some things it's better than other things. And it's fully integrated, so it's basically, I don't think you can really achieve this at all with third party solutions, because they, as impressive as they are, you're never gonna be able, I believe to, at least not in the foreseeable future, to integrate them as deeply across the board, and to tailor them specifically to what you want. And by the way, you do then you're, I don't think you can, but even if you could, then you'll be locked in with these vendors, which is dangerous because you'd be completely exposed to massive increases in prices potentially in the future. So partly with Allspooner, we achieved, I think, much better effectiveness because it can do a lot for you in a very efficient way. And I'll give you an example in a moment. And partly, we achieved a good separation and very low levels of dependency from any provider of AI infrastructure or AI models. Examples of what Old Spooner can do, it can do data analysis for you. I was looking last night, I was looking at the results of certain AB tests on StreamYard, is one of our products, and normally I would have had to ask someone from the team, a data scientist to pull the data, prepare the analysis, or at least do some of the work for me.
00:54:34 · Luca Ferrari
I can do my own data analysis, but I would have needed someone to do some of it, and certainly would have taken either me or this person multiple hours. And I simply chatted with this, with my old spooner and I told it, oh, please, you know, go on and do this or that. And maybe five minutes later, had the analysis. I was actually very positively impressed. It gave me plots with highlights, very detail oriented and nice. Or if you're running one of our businesses, can, and I saw this firsthand multiple times, you can go to your old Spooner and you can tell it that you saw a bug. This happened, it's a real thing that happened with Evernote. The general manager of Evernote told her old Spooner that she had encountered a bug as well, she was using the app and tasked the agent to check on our customer support platform that whether the bug was widespread among users or her report was the only one, and then to go into the code base, identify the root cause, code a fix and ping the engineering lead for that project to review the fix and push it to production. So maybe she spent, I don't know, three minutes to provide these instructions, and then I'm sure the engineering lead had to spend maybe an hour or two reviewing the code, but the bug got fixed. And this is something the same day, and this is something would have taken weeks probably between back and forth between different people and way in a lot of hours of actual human work to get it done. So thanks to this and many other technologies we've built were probably two, three times as productive easily, and the list is long, but this is roughly what to expect. Now potentially from a business perspective, even more interestingly, AltSpooner really helps us, as I said, stay independent of providers. We have built our own orchestration, so every time you task your agent with something, under the hood, we have an algorithm that will select the most appropriate AI model, or multiple AI models to get the job done, considering, say, quality effectiveness, but also cost. And it can draw from many, many different models.
00:56:58 · Luca Ferrari, S1
We end up using open weight models, we self host, So these are basically free, I mean, there is a little bit of cloud cost, but for 99% of the requests and tokens, and the frontier models generally closed weight through APIs for, I don't know, maybe 1% of the requests only for the most complex tasks or for supervision. So they will, sometimes we use them again, automatically to check the work that got done by the slightly less intelligent models as a more senior engineer would with a more junior engineer, but that's way cheaper than actually doing the work, which is often perfectly fine. And even if it's not, the smarter model will provide a few pointers and then the less smart model will go and fix it. So because of this, we've been able to stay independent of any one vendor and keep our costs for tokens super low, basically negligible Wow. At our [00:57:57] S1: When did you start deploying these AI agents? Also your point on not having a third party is very counterintuitive to all the marketing that's going on right now with all these AI assistants and applications that are coming out.
00:57:57 · S1, Luca Ferrari
[00:57:57] S1: When did you start deploying these AI agents? Also your point on not having a third party is very counterintuitive to all the marketing that's going on right now with all these AI assistants and applications that are coming out. [00:58:11] S0: We work with all the big labs, they offer great products, we use them enthusiastically, I'd like to think we're a good customer, but we don't want to be dependent on any of those specifically, if it can be avoided. And there are solutions out there that are much cheaper and often deliver essentially the same quality, but it takes, I think, strong engineering capabilities to, and the right culture to be able to harness those possibilities. Obviously the easy approach is hand over the keys to one of these companies and buy their more expensive product. It will make you come across as AI enabled faster and more easily, but it will be, I would say much less effective, but certainly way more expensive, literally orders of magnitude more expensive. [00:59:02] S1: I don't know if you saw this, but Jensen just declared we've reached AGI with OpenAI's Astra model.
00:59:09 · Luca Ferrari
[00:59:09] S0: Some people say that, maybe true. I mean, I don't even know. I've heard different definitions of AGI and none of those are super unambiguous. I don't know, nobody can tell. But it's obvious that it's very smart, let's put it that way, whether we should call it AGI or not, I'm not sure, nor do I care too much to be honest, I don't think that AGI, even if we can describe it very specifically as a threshold, don't know that crossing it is a particularly, like the moment we cross it is a particularly noteworthy milestone. You could easily, you know, people sometimes define a GI as AI can do everything that any human can do, but better. I don't know that that's necessarily more exciting or scary than AI can do. 95% of the things that humans can do, AI can do just as well or better, but 5% it can't yet. I think this is almost as exciting and almost as scary, you know, depending on what the 5% that it can do is, but assuming it's not handpicked to be the things that keep us in control, but just a random set of tasks, we happen to be, our brains happen to be more capable at, so I'm not too keen on whether we pass the AGI thresholds or not, but the trajectory is clearly one where AI does things better than us, increasingly so, both vertically, so the gap in how much better it can do a certain task is growing unsurprisingly, but also horizontally, like the percentage of tasks it can take on and do better than humans is increasing. I don't think there's any stopping that short of some sort of world war we regress to the middle ages.
01:01:01 · S1, Luca Ferrari
[01:01:01] S1: Well, seems like now the next benchmark is RSI, and that with it comes a lot of fear mongering for cybersecurity and cyber risks. [01:01:09] S0: Yeah, it's a very fair fear. I'm equal parts enthusiastic about AI and absolutely scared shitless. Think this could be, it could definitely make a case for this being the greatest boon for humanity ever by orders of magnitude. Possible? Yes. Plausible? Maybe. I'm not sure. Equally, it could be the thing that out wipes or creates equally awful scenarios. What's like, I don't think humans of course want the latter. The point is, we control it? Even before it's so powerful that controlling it is all that matters. Think we, while we are approaching that point, maybe we're not quite there yet. But regardless before that, can we prevent it from falling into the hands of, like, some sort of degenerate or evil person? Because, you know, similar to nuke... Nuclear weapons, but potentially worse, I think, because... I mean, I'm no expert in nuclear weapons, but I think number one, it's very difficult to do irreversible widespread massive damage with nuclear weapons, such as almost wiping out humanity without killing yourself in the process. You could use AI to your benefit while causing immense damage to everybody else. I think it's easier because it's much more surgical. Can we do that? I'm not sure, it's pretty scary. [01:02:46] S1: The accessibility point, think, is the most alarming because it's literally accessible. The barrier to entry is so low.