[00:00:00] Michael: Hello and welcome to Postgres Om a weekly share about all things PostGresQ. I am Michael, founder of PG Mustard. This is my co-host Nikola, founder of Postgres ai. Hey, Nikolai, what do you want to talk about this week?

[00:00:11] Nikolay: Hi, Michael. It's very boring. Next time I should, uh, uh, introduce ourselves. Next time it's my turn. Like, let's have less boring episode. let's talk about

It's time. I guess during last three months, uh, since G P T four released, uh, everyone already tried it at least a few times

[00:00:30] Michael: Yeah, exactly. We're not gonna, hopefully, we're not gonna cover any of the basics in terms of what it is and everything. I think most, uh, pretty much everybody knows that already, and if not, I'm sure other people have done it better than we possibly could. But you want to talk about how people can be using it with Postgres, whether they should, what the risks are, what the opportunities are, that kind of thing.

[00:00:49] Nikolay: all right. let's talk not about, PPG vector and how to put embeddings and how to extend behavior of child and so on. But, uh, GT four and so on, like JT 3 45. Let's talk about, user experience to support, SQL developer experience and, you are database engineer or you are backend engineer who needs to work with pogs.

Maybe you are DBA who needs to work with pogs. What can it bring you pros and cons. Let's talk about these aspects. I, I, I mean, I mean, PG Vector is a great thing and we see Super Bays new and even rgs recently started to have to provide it. it's super cool. A lot of, new ideas are being implemented, but let's just put this topic aside and focus only on regular user experience.

I'm a user, I mean, I'm a developer. I'm, I'm developing my app, or I'm dba or I'm writing SQL code, or I am analyzing my data using SQL data analysts or something. how can help, uh, and so on.

[00:01:53] Michael: Sounds good. I like this topic cause I think we're gonna, it's gonna be one of the ones we disagree most on. So I think I'm, uh, sorry in advance,

[00:02:00] Nikolay: I hope finally,

sorry for interrupting. I hope we will disagree a hundred percent, uh, in all aspects on this topic.

[00:02:06] Michael: Yeah, maybe not all aspects, but I think I'm a lot more negative about some of these things. Or a lot more conservative, maybe cautious, than you are based on our previous discussions. Yeah. Person maybe, I don't know.

[00:02:16] Nikolay: You are not alone. Let's see. I take it as a, as a valid assumption that everyone already tried it at least once. The only thing is that is, uh, GT four version steel for paid customers, or it's for free as well,

[00:02:32] Michael: I believe it's still paid only.

[00:02:34] Nikolay: right? default this 3.5, A lot of people used 3.5 and my first point, don't use it.

[00:02:40] Michael: Yeah, version four is a lot better in some very

[00:02:44] Nikolay: Just don't use three, five. Don't use three five. Don't use integration with three, five. Don't use it. Forget about it.

[00:02:50] Michael: So quick thing if you see people sharing screenshots, I think an easy way of telling the difference is on version four, you'll see there's a, it's a black logo, uh, when chat GPTs responding. And in 3.5 it's a green logo. So that's an easy way of seeing like what somebody's used

share

[00:03:07] Nikolay: in all, in all my experiments, 0.5 was quite silly, stupid, and, very basic and a lot of hallucination. Four. Sometimes a lot of hallucination, steel, but so more like deep and wide. so just use version four. It's much, much, much better. That's why, when it was released, a new wave of attention, has started. But, as I said, you're not alone being pessimistic, assuming everyone tried, right? I'm checking, poll results. I, a couple of days ago on LinkedIn, on Twitter.

I ask the, do you use chat for your post related work? Three options, not at all. A few times per week or already A daily user. And the Twitter, 80%, not at all. 80, 15, few times per week.

[00:03:57] Michael: How many voted?

[00:03:59] Nikolay: How many vote at, of 99?

[00:04:02] Michael: Okay, well,

[00:04:03] Nikolay: not many, but I started it, two days ago, so sometimes I have more votes during two days and already a daily user, just 5%.

Just 5%. I'm a, I'm close to being already daily user. and LinkedIn. Similar, actually better in terms of PO positive, usage, not at all. Uh, 60 few times. 30 daily user. 10,

[00:04:28] Michael: And sample again.

[00:04:30] Nikolay: sample 81. Also not

many.

[00:04:32] Michael: Yeah. But not 10.

[00:04:35] Nikolay: Not 10. Yeah, around hundred. But, not many. And, with thousands of oppressions, I guess many people just not vote,

[00:04:41] Michael: Did that surprise you?

[00:04:43] Nikolay: what I feel, when GT four was released, huge wave started it was boom. Right now we are on declining side, but it's going to change I bet.

because of course, G P T five, when it'll be released, uh, later this year, maybe next year,

well,

it'll be interesting. there are some rumors, I guess only, uh, maybe later this week, maybe next year. But, of course this will be boom. But also I think, are developing, new integrations.

they're adjusting methodologists and so on. Uh, it's, going to, uh, stay. In my opinion. I'm very positive that it, it's, it's a super helpful thing and I'm using it almost daily. but, many people disagree. I see it. We see posts like, don't use it at all, which is wrong. I say use it a lot.

Just use version four. Of course, it has like this stupid limitation, 25 messages per three hours. I often bump into it. And then it says, okay, switch to 3.5 and 3.5 is just terrible in my, in my experience compared

to what four version four provides.

[00:05:46] Michael: So what, how come you're using it so much? What kinds of things are you using it for?

[00:05:50] Nikolay: Good. let's explore this, where to start, uh, develop or administrative, like infrastructure tasks,

[00:05:56] Michael: maybe focus more on what's the most common things you use it for? Like in terms of frequency, how come it's nearly daily?

[00:06:03] Nikolay: I don't need, help with sql usually.

Well, let me to be honest, yesterday I saw I don't use, uh, procedures often and yesterday I needed to code some procedure, not function well. I could do function as usual, but I wanted, transaction control, so, It's something which talks to external service right from PO's primary, which is not good.

But, in this case we do it, understanding problems. So, I saw some examples somewhere in documentation and saw Begin atomic. And I actually didn't know what beginning Atomic is. Well, I can guess, right? So I asked Chad, d p t, it's, says, I'm like, version, September, 2021.

Outdated, posts doesn't support Begin atomic. It says, obviously it does just in new versions, it but then it says, you know, in SQL Standard, this It means that, the whole body will be single transaction. Well, okay. Obviously, and I'm sure a positive documentation would explain it just not on the same page.

I, just searched for it, didn't find it. Chat was faster away to understand it. Well, I understand. Okay, now it's single transaction. I don't need that. I need a regular beginning when commits, to split to multiple transactions. was useful. I mean, it was faster way to, rather than Google or, uh, trying to find it in documentation.

Documentation has this on different page obviously, right.

[00:07:22] Michael: but can I just add at this point, and it's super interesting that you use it for that. I wasn't expecting kind of factual checking to be number, like the first use case because my main fear with using it for things like that later, if I, if I check the source code, even if it takes me a few seconds later, if it, if I check the Postgres docs, I trust that it's true what it says.

If I check jack, g b t, how do I know it's true? Like, of course with a, with an example like that you believe it's true because that's also your instinct as to what begin Atomic was going to mean. But I get incorrect or false results often enough that I now no longer trust what

[00:07:58] Nikolay: I don't see any problems here at all. let's, um, have some spectrum of, level of trust. Documentation probably is the top trustworthy source of truth, but it has bugs sometimes.

Also True.

[00:08:12] Michael: Yep.

[00:08:13] Nikolay: Just, uh, it's not often at all. It's not ideal. Still, it might have false statements sometimes or misleading, but probability is tiny.

[00:08:23] Michael: Especially tiny for things that matter, right? Like if it was something that mattered, it would've been more likely to have been corrected, like right. It's more likely to have bugs in places that don't matter so much. That would be my, instinct.

[00:08:36] Nikolay: Right. Now, let's, take books for example.

[00:08:39] Michael: Mm-hmm.

[00:08:40] Nikolay: Give me any book. I will find a problem with logic and, false statements. I mean, PO's book, give me any book. I will find it because they're much lower in terms of quality, I didn't see any super quality books, uh, at all.

[00:08:52] Michael: the art of Postgres, dimitri.

[00:08:55] Nikolay: I must admit, didn't read it fully, but, okay. This is my homework. I will find problems in it.

[00:09:02] Michael: Sorry, Dimitri.

[00:09:03] Nikolay: Ch challenge accepted,

[00:09:05] Michael: Yeah,

it's

a very

good book.

[00:09:07] Nikolay: yeah. well, maybe it'll be, not with logic, but something like, controversial statement or something easy. I will find it.

[00:09:14] Michael: What about Marcus Winn and Sequel? Performance explained.

[00:09:17] Nikolay: Very good. Very good example. Uh, yeah, this is very good example. Second challenge, except that I will find some issues there

[00:09:25] Michael: read that one? Surely.

[00:09:27] Nikolay: Well, uh, I read some parts of it,

[00:09:29] Michael: Okay.

Also, very, good book

[00:09:31] Nikolay: this is very quality book. You, you chose good examples.

[00:09:35] Michael: Yeah, because, and this is my point about Google search as well, right? Like, so documentation, very high post, so Postgres documentation, very good books. because I know the authors, I have a higher level of trust. So if,

[00:09:47] Nikolay: But there's just couple of people plus more couple of people who are corrected. It's just tiny group of people. They for sure make some mistakes. For

[00:09:55] Michael: I am not question. I'm not questioning that. I'm just saying it's maybe slightly lower than documentation because I don't believe as many eyes have gone on it and I don't believe it's as easy to correct. Although Postgres documentation could be easier to correct for sure. Um, Then Google, like search results for example.

Again, you get some metadata, you get some information about who wrote this, when did they write this? You can like check some sources and you, you build up once, like over time, even on Stack Overflow for example. Over time you start to notice some people that give amazing answers and some names. I just trust more than others when, when it comes to like a stack overflow response.

But on chat, G P T. I get no information about the source. where did this information come from? wh what's it derived from? You get No kind of, I, I don't get any signals as to whether to

trust it or not.

[00:10:43] Nikolay: you see Marco Swin, perfect author. I like have huge respect to

this author. But, if you don't know this name, where is trust coming from?

[00:10:53] Michael: Yeah, I understand trust is built over time, for sure. the first time I read it, I check it's true. Okay. Yes, that was, that was good. And then the second time, and then the third time and the, and the 10th time. And sometimes they teach me something that I didn't expect or they question an assumption I had that I didn't realize I had.

And then when I verify it's true

[00:11:11] Nikolay: Verify is the key. Thing I wanted in our, in this episode to deliver.

[00:11:17] Michael: Yeah.

[00:11:18] Nikolay: I don't trust anyone. It's sometimes I don't trust the commutation either. And in my whole practice with POGS consulting practice my daily like development and so on, I often don't trust myself. like, maybe I'm wrong and this is normal.

And the key point you need to verify. You need to experiment, test, and so on. You need also, you need to know how to proper do it. For example, if you do something with, large dataset, use large dataset and so on, And same here. of course some after produce super high quality, statements and, share super high quality thoughts.

For example, Tom Lane in.

[00:11:56] Michael: Yeah.

[00:11:58] Nikolay: in 99% unbeatable logic, right? But sometimes some statements, even from Tom Lane, I could argue and some people could argue as well, because, uh, life is not like simple, like maybe for example, Tom Lang said something, But for you, for your particular case, it's not, cannot be applied, for example.

Or you have different experience and, Tom Lane said, don't use this approach, but you're still using it and you find it super helpful. In your case it might happen. so if you consider charity PT as some consultant, and version four should be used again Assume everything with grain of salt. Verify, the magic is that you can tell, you know, you say something, but there is, there is, uh, something wrong here, right? And it, uh, can, uh, fix problems. It adjusts. So this chain of communication, sometimes it's annoying. Like, every answer starts. I apologize for confusion, right?

I have it so many times, but it's super helpful. Like, same like, I hired someone. and the guy is super fast in answering and has super wide, knowledge, Sometimes quite deep as well, but might, hallucinating sometimes it's okay, and if I have a very good way to verify ideas and iterate, I cannot understand how to live without it at all.

[00:13:19] Michael: Yeah. So this is, so you're saying that having access to chat G p t four is like hiring somebody with incredible breadth of knowledge, sometimes very deep knowledge, and they're, always online. They always reply straight away. they're never in the middle of something else or except, you know, when, when you run out of queries per hour.

Yeah, I completely agree with that. The one thing I would say is a bit

[00:13:39] Nikolay: No, no, no, no, no,

You hired someone and, you don't trust, uh, this employee hundred percent. Uh, you need to be five ideas and sometimes tell you no, you're, you're probably wrong. You can work with, with tragedy as with. Person and interesting that, in my consulting career with PO Consulting since I moved to California 10 years ago, I was no name here, right?

Initially, and when, when I go to companies like Shwe, GitLab and others, they doubt everything you say everything. So if you say something, uh, you need to prove you are right the best way to prove, obviously is to try and show how it works and.

Explain in very detail how to repeat it so you say something, you claim something and you show how to check it, to verify it. And if you consider child is not a hundred percent reliable thing, it's not documentation, right? It's some consultant who. Is wrong sometimes, quite often wrong, but we allow it to be wrong sometimes.

Right? But we need the, like, without a way to verify it, you will end up, concluding that, uh, this is a bad tool with verification approach. doubt everything and verify it.

And if it's, shows me mistakes, show these mistakes and ask to, uh, fix them,

it works very

well.

[00:15:09] Michael: one Big difference versus a consultant or human in general, or at least the, the types of humans I would employ is that When it is wrong, it's very confidently wrong. it doesn't say, I'm not sure, or I'm, yeah, it doesn't ever, it's very, very confident in its answers, which is very different from humans.

so that is, that's a big difference, but it's also a lot cheaper than humans. So I think it's like, it's not a fair comparison

at

[00:15:37] Nikolay: good consultants always, put some, doubt element into all statements like, I would say that or, most likely blah, blah, blah and so on. Like, there are many, many intro phrases consultants use, which, assume some probability of mistake. and my own style is, I always say, I'm not sure.

I never, I'm never sure, in our own episodes, I was wrong many times. For example, I just commented, I responded to some YouTube comment where I corrected myself I claimed delete has a index amplification problem. It doesn't because delete just puts to, to, topple indexes are not touched.

And I corrected myself in YouTube comment and someone asked to explain, to elaborate, which I did. I can elaborate some other time, but when I elaborated in that comment, I thought maybe I'm wrong. You know, I checked dogs, I checked, I didn't check source code. I checked, got OL's, uh, pogs.

Inside, inside pogs, uh, book,

[00:16:38] Michael: Yep.

[00:16:39] Nikolay: no, it's not in journals. Inside pogs, in Russian. It's inside pogs. Perfect book, by the way. translated to English and so on, but

[00:16:47] Michael: Perfect book without mistakes. Is that what you're saying?

[00:16:50] Nikolay: Well, I go, yeah, yeah, yeah. Well, I, I'm sure there are mistakes there, but Yeah. Well, if you ask, I, I'm sure if you ask, I go, he will admit some probability of mistakes

there. Good ath, good authors always admit and good consultants always. So what I did eventually, before I put comment to YouTube, I went to P SQL and verified everything I'm going to say. And then I added it to the comment and I found in my consultant's practice that we always need to do, if we claim something, show it.

[00:17:23] Michael: And that's something, chat b t can't do.

[00:17:25] Nikolay: it can, I mean yes and no but you can assume it can be wrong. Assume it can be hallucinating. Ask Chad Deputy to, to provide some ideas. Example, it's a brainstorm tool. It's not verification

[00:17:41] Michael: There you go.

[00:17:42] Nikolay: but then you need verification tool. And I'm going to shamelessly advertise what we do. This is perfect Verification to database branching thin cloning.

If you have charge pity and if you have cheap and fast, way to check anything, well, of course you cannot check hard things like, hardware change and so on, but SQL query behavior on large data sets. Think loading at the top is branching is priceless here. people work on bottom list on serverless, but we work on priceless, which is database branching through database branching on any, any location.

this is my company, my product. So you take share PT and you go verify on the clone. Clone is provisioned in the second for free? I mean, you need to prepare it, but then you have a zero extra cost and then you can iterate. Okay. I say, you know what? On my one terabyte, Database. This is what I see. And I have, like, this is, I, I'm considered like, as a human. Like it may, it might, it might be mistaken. So you propose something which is not working on my one terabyte database, and it'll adjust. And in, in a couple of iterations, you'll have much better results. And I think this is something you hear, with code, it works pretty well.

I already used it for Ansible coding. show scripting like some advanced short scripting, react coding. I'm not a react expert at all, but I did some many things. I wow. With Charlie P I can do many things because with code you can easily check it very quickly because you can rather, uh, if it's c probably you need to, to wait until it's been compiled, right?

[00:19:23] Michael: That's a really good point though. It's easy to verify quite quickly.

[00:19:27] Nikolay: with databases, you need database branching. So go to noon. If you cannot go to noon, go to us. That's it.

[00:19:35] Michael: but, so, okay, so we've actually skipped a bit. We've, gone, originally we were talking about kind of verifying things that you could check in the docs, for example, or asking it for factor information. Now we're talking about trying out ideas like idea generation, brainstorming. and now I get for simple, for isolated things.

it's very easy to test those things. What about system-wide is issue, like if you're asking about, I don't know, some modeling question would you use it for that kind of thing where it's like you or maybe asking architectural decisions, is this a good idea or is that a good idea?

[00:20:05] Nikolay: it's a, everything is a good idea. It, absorbed a lot of stuff. It absorbed stake overflow. Its Of course, sometimes it's, again, hallucinating. Easily. But if you prepared all responses, like take it with grain of salt. This is what I think as would any good consultant say?

without ion, I can say this, but it should like this needs. Thorough verification, but from top of my mind, I think this, if you prepare something like consultant style, phrase to any GT responses, if you use GT four, it'll become very good, useful tool.

You unlock your ideas, and then you go to verify them. Then you come back and say, this is not working, and so on. Data modeling. Perfect example. seeding is perfect example. I have, uh, new tables. I, I haven't launched yet. use it recently. I have some roles already there, which is probably represent my data in future.

Can you generate more? Okay. I have 10 rows. I want 10 gigabytes of rows. Can you help me? It says, okay, use this snippet with generate Sirius, this is my real life example. Charity PT version four provided some pretty interesting, generated serious example with very useful comments.

You can adjust these parameters in this snippet

[00:21:26] Michael: Mm-hmm.

[00:21:26] Nikolay: was super helpful. I would spend some time generating this. I used it and said, you know, I wanted 10 gigabytes, but this produced only three. It says, no worries adjusted. We added seven more gigabytes in second iteration, I have now some data set.

I can start testing my application using this data set. This is very valid, uh, case.

[00:21:49] Michael: Yeah, I can see that I've had similar experience with helping me generate, I needed an example for a blog post, and I think sometimes I get blocked on those, you know, just trying to think a little bit creatively about

[00:22:01] Nikolay: So you also use it,

[00:22:03] Michael: Yeah, I've used it, but my experience is different.

I u and I've, I pay for four as well at the moment. I'm, and

then,

[00:22:10] Nikolay: to stop. Hold on.

[00:22:11] Michael: Maybe gonna stop. I haven't used it for a couple of weeks. but for sometimes, like if I'm blocked creatively, I've found it helpful. I th I like your, premise that it's, it can be used as a brainstorming tool. I very much agree, that it can be really helpful for some creative ideas,

[00:22:26] Nikolay: In brainstorm approach, like enterprise style, we always say like, during brainstorm, don't criticize ideas. We just collect ideas.

Don't demotivate people, so don't demotivate charity. It, it, it tries, it's it's best, you know,

[00:22:43] Michael: Well, but it also lies, and that's my, like, it also doesn't

say, I'm not sure or it depends. Yeah. But it, it lies way more explicitly. And I, and one thing I'd say is The reason lots of people disagree on this is like, I think it's a little bit half glass, full half glass, empty type, viewing of it.

When I see people praising chap ptt, they could easily get an answer that has 10 bullet points in it, and seven of them are quite impressive and three of them are garbage. The glass half full person's like, oh my goodness, how did it get those seven?

So good. And they know enough that the three are not dangerous or are not, a problem if it's talking complete nonsense or if it's suggesting really bad ideas. Whereas I think it's a very fair criticism to say. The, there's, there are three quite dangerous pieces of advice in there. Please don't listen to what this thing's saying.

And I think that's where some people are coming from. Like, I think we saw that post by Christoph Pets just, a couple of days ago saying, don't use it for advice because sometimes it's really badly wrong and some,

[00:23:43] Nikolay: naive. This, that's naive, statement.

Uh, everyone will be going to use it.

[00:23:47] Michael: Do you see the different like angles? You could be viewing the exact same 10 bullet points and one person says, this is amazing, and one person says, this is

[00:23:54] Nikolay: If you hire a consult consultant. It, it's going at some point to be mistaken. Again, prepared my perfect store responses like, if open the eye would be. More like, I don't know, they should probably do it

pre instead of, I apologize. for confusion, they should prepare in the first answer, they should prepare.

Like, take this with grain of salt. This is what I think without

verification. Go and

verify it.

[00:24:24] Michael: Yeah. It would be

much less impressive.

[00:24:27] Nikolay: so in 2016, 17, I started to. To give talks about database experience, like my, approach is verify everything.

And of course it might take ages to verify everything. That's why we need to simplify it, to speed it up to, to make it cheap.

Experience. And we found a way database branching. it's recipe for most of needs that backend engineers have, not DBAs and infrastructure teams. They have different needs, but, programmers, backend engineers and front engineers, sometimes we can satisfy them very well with, thinking and database branching.

So with this approach like. zero trust approach. If security have a zero trust approach, why, couldn't we have this with development? In development? Okay, someone, uh, has some idea, but without verification, we have zero trust into it. We say thank you, but we are going to check it with this approach.

I could not see how you can consider charging PT as a bad tool at all.

[00:25:34] Michael: Right. I think we've covered accuracy quite well. I wanted to ask a couple more things. First one, privacy. like, how are you feeling about their privacy policy in general? yeah, how do you think about that

[00:25:47] Nikolay: Okay. This best,

tough question before We go there. let me remind our listeners that we had beautiful session with Charlie PT and SQL Optimization. It's on Postpositive tv. YouTube channel. I would like to have another one. It would be

great,

[00:26:03] Michael: There. were some funny moments and some, Yeah.

it was a good, I think we

spent a long time in the end.

[00:26:07] Nikolay: Yes, but, summary was, uh, it exceeded my expectations,

[00:26:11] Michael: Yeah, it was quite good. A few things and, and I think we gave it quite a good example. We only, I think we only checked one

[00:26:16] Nikolay: it exceeded not at first iteration.

So, uh, secret sauce was, Verification because I said, you know, I tried this, it doesn't help. But I also said, I know there is a good, uh, answer. Well, you can say it always right. So

it's not something

[00:26:34] Michael: to, right?

[00:26:35] Nikolay: I didn't, reveal anything.

it found proper answer after iterations and my like, I insisted, right? So, so

[00:26:43] Michael: Extra prompts like first,

[00:26:45] Nikolay: extra prompts.

I said, let's try again, let's try again, let's try again. And eventually it provide a very good example and we, we find it. And that index was great. And this is what I also explored myself also not at first saturation.

No. Like there are no perfect humans. So, if you consider charity as a hu more like a human, it helps.

[00:27:08] Michael: Well, on the topic of SQL optimization, I obviously as a product maker, I was thinking, should I be using this in my product or like how could I, is it a threat? You know, it's a direct, in a way, if it can be used for SQL optimization. It's a direct competitor to the product I make.

So it's naturally a very serious thought and maybe, maybe I'm deliberately negative, you know, consultants being negative on it and telling you not to use it could also be self-preservation, right? It's same with a tool maker who it could be competitive to. but what I wanted to say was the reason I've not.

Used it yet in the product the reason I'm not planning to anytime soon is we want to give advice that is trustworthy. That does come with a lot of, uh, you know, does optimize to not give those false positive, those, uh, those bad advice that's dangerous. And that

like, that's something I'm not willing to Exactly.

Yeah, so I think there's an interesting trade off there.

[00:28:00] Nikolay: My approach is very different. I've learned in my life that, it's good in terms of finding interesting results. It's good to generate a lot like this brainstorm approach. You, Jeanette, crazy ideas, a lot of them. And then just go and verify them. If verification takes, limited time and limited budget, it's great.

So I would rather choose 10 ideas. Three of them are crazy and wrong. Then just ideas, which are, are trustworthy. I would, I would grab 10 and go verify them. My approach is like, give me all ideas. I'm going to like, I have zero trust to anyone, any statements, so I'm going to verify myself anyway. And without this verification component, we have only half story here.

[00:28:48] Michael: let's skip. to, privacy. And I want to, I want to ask you about something else as

[00:28:52] Nikolay: Why, why do you say privacy? Okay, sorry.

[00:28:55] Michael: This is same thing.

[00:28:56] Nikolay: So privacy, I, it's not, it's not resolved problem. And I remember in the Beaver, they have pulse statement, you can disabled here, fully disabled here. Integration with charter P, which is of course demonstrates that they know that many people

are concerned. And we know some countries haven't banned. Some already un banned. So like it's still shaky area, right? It's not, fully clear how to approach this because my question to you like open ai, is it like about openness of everything?

If you talk to Charlie PT reviewing some data and some secrets, is it go, is it going to be open as well? Because it's open ai.

So now my, my, SSN bank account routing number, everything is going to be open or what?

[00:29:45] Michael: They say no, right? Like, uh, I know you're joking, but they say no. But do you tr like it's an American company, right? Like it's for-profit company now? I believe, like I know it started as a, I know it's supposed to be more of a non-profit initiative and, but now I believe it's a fully for-profit company.

[00:30:00] Nikolay: I pay

[00:30:01] Michael: Um, Yeah. Well, yeah, you could pay a nonprofit as well. but yeah, so I, I think there's a real concern there. I've definitely seen some companies also not allow their employees to use it for these,

[00:30:12] Nikolay: Yeah.

[00:30:12] Michael: these concerns.

[00:30:13] Nikolay: burn.

[00:30:14] Michael: But the main reason I see people not using it or being told they can't use it is actually around.

IP So like, in terms of using what it tells you, like about, about, oh, I guess it's around the ethics of it, how did they, you know, the gathering of that information, do you have any concerns on that side of, of

how they got the information?

[00:30:35] Nikolay: Let's put, it into parts. we have query SQL queries. We have explained plans, and we have some additional parts of ip, like ideas and, uh, like approaches and so on. I'm strong against patterns and ideas. Ideas should flow. Like, uh, like scientific approach. You go to conference share ideas, everyone takes it, improves it, and so on.

And somebody, somebody maybe not you will win. Of course, many people will not, agree with me in terms of patterns and ideas, but this is my, opinion. think it's bad for innovation and for progress, but if you talk about queries and plans, do you think, what's the best name for a project Will that will solve it? PG mask or PG gloves. Gloves or something else.

[00:31:24] Michael: I don't understand.

[00:31:25] Nikolay: Okay. Uh, imagine you talk to t and you show some query and some maybe some plans or execution.

Explain plans, which, uh, Peter Masters is working with for as well. And before you're send to jt, you just transform all table names, object names, and parameters to some something,

[00:31:46] Michael: Oh, like a, yeah. Okay. So like Deez. Deez does that, right? Like anonymization tool,

[00:31:53] Nikolay: yeah, has it for plants expand plan, imagine this project will transform all. Object names to something

[00:32:01] Michael: So you're saying you.

You can anonymize it, send it through, get it back anonymized,

re like de

[00:32:07] Nikolay: Right.

Right. Right. Or mask and mask or anonymize. The anonymize

you keep, mapping. I think it's good.

[00:32:14] Michael: sure. But casual users, aren't going to, like,

if we're talking about using chatt PT directly at the moment it's not, that's not super practical. And, and if it's a time saving tool what's the point anymore? But yeah. So, but what about the eth? Like you don't have any concerns around it,

[00:32:28] Nikolay: I have concerns like, my query can have emails. I don't blame them to be leaked or SN numbers or some medical info. I I don't use data, to be leaked, so of course I need some gloves or masque. What's the best name for this project? Ma Gloves or something else?

[00:32:45] Michael: pgn non or something. Maybe that's not good. Good

connotations.

[00:32:48] Nikolay: Okay. I'll think about it more, but this is, uh, I, I, I bootstrapped it slightly, but I think it will. Have some negative effects because when you talk to, if you have some experience, yes, I, you say this is stable persons, it has a calm email and from names, CHATT also can suggest something useful.

For example, it can ha, it can suggest you to have index lower on email, for example. To have it cancer sensitive or to use, data type ci text, cancer sensitive text. Right. If you convert it to column one, it'll, it won't, like, it'll work less sufficiently. Right. But you, well, you lose something here.

Some senses right from that can be derived from. Names, but at least you are protected if you transform to some weird names using some tools or maybe your own new ones. And then, uh, in responses, you transform substance to back to it. It'll protect your, your you almost fully. Some people told me that some SQL queries, their structure can be also like considered as ip.

Well, I can share very advanced SQL examples. Very advanced. I learned from other people. This is about area three patterns. Okay. Like

[00:34:13] Michael: This,

[00:34:15] Nikolay: explain

[00:34:15] Michael: it is not just patents. That It's not just patents though, right? It's also Even coming up with those, it is somewhat, might be somewhat novel and the idea that

like

[00:34:26] Nikolay: Here. I, I, think everything should be open. All the recipes, pog, this is open. Recipes should be open. Talks should be open. Recordings should be open. Everything should be open

[00:34:38] Michael: Nice.

we've run long. I think it's a good one still,

[00:34:41] Nikolay: It was supposed to be very short episode.

[00:34:44] Michael: yeah.

how do you wanna conclude?

[00:34:47] Nikolay: Vfi check

test experiment.

This is the key.

[00:34:51] Michael: like the thinking of using it as like a brainstorming tool, like, as a

creativity tool, give you ideas and very much don't trust anything it says,

but you might get some good

[00:35:02] Nikolay: Same as humans. Don't trust our

episodes.

[00:35:05] Michael: Yeah. Nice one. Well, thank you, Nikolai.

[00:35:07] Nikolay: Okay. Thank you Michael. Good. See you.