Godel continues to establish an AI Community to harness the rapid growth of AI technologies and support Godel’s partners with most up-to-date solutions.

In this episode of The Godel POD, our host Kara Goward, PR Manager at Godel is joined by Godels Head of .NET Andrei Salanoi to discuss his journey of AI, how implementing AI can be beneficial for businesses as well as debunking any AI myths.

This is an edited transcript, for more conversations on the latest in tech, subscribe to The Godel POD on Podbean and Spotify. Podcast jingle provided by the Hideout YouthZone.

Podcast transcript

Kara Goward: Hello and welcome back to The Godel POD. My name is Kara Goward and I’m PR Manager at Godel. Today I am joined by Andrei Salanoi who’s Head of .NET. He’s here to discuss all things AI and Godel’s growing AI community.

So Andrei, would you like to give an intro about yourself and your role at Godol?

Andrei Salanoi: My name is Andrei Salanoi. I’m Head of .NET division and also Head of AI community and have been with Godel for almost 10 years. I’ve passed through all the positions from junior developer to lead of architects and actually to managerial which is Head of .NET division. The biggest one we have at Godel.

Kara Goward: Cool. And how did you get into AI at Godel? I know it’s a huge topic at the moment, lots of predictions out there. So I suppose what made you get into AI?

Andrei Salanoi: Interesting story. Usually when something creates a lot of buzz around, I tried to like study this topic and AI was not an exception. The previous topic was Docker and .NET core. Now I spotted that something is happening in the AI area and I started to investigate and it looks to me interesting and I get an impression that is going to have a big impact. Along the way, we actually built the software and it was interesting for me personally. As a coincidence, I got an offer from Viktar to drive this initiative, Viktar Barysevich at Godel and I thought, it’s in line with my personal interest and this is how I became head of AI community.

Kara Goward: And what is it about AI that interests you?

Andrei Salanoi: Funny fact is that when I was a student, the speciality is was studying at college was called artificial intelligence. But at that time I have no clue why it’s called that but you had relevant disciplines related to AI. And for like almost 20 years, there was no use for this knowledge. And now, like a coincidence, I started to be involved in AI like in the real world with real-world usage scenarios so.

Andrei Salanoi: In a nutshell, it’s a way to do more with less effort. Maybe that’s why it’s a nice idea that the programme will do everything for you and you will get money for this.

Kara Goward: And then this brings down to a really good question. There’s a lot of myths about AI, so around changing people’s roles, maybe in coding, design, writing, what myths can you debunk about AI?

Andrei Salanoi: I would say that maybe fear that once AI is involved, everyone will lose their work which to my mind is not true. I would say that. AI should consider it as a way to be more productive so I could not believe, at least for now, that AI could actually replace a person on creative work. What it could do. It actually saves this person’s time onboarding repetitive staff and simply speaking, this person may become like 10 times more productive in doing the stuff the person actually doing manually before. Now that AI is under control, this person can actually do this stuff and this person is to generate ideas and supervise the outcome from this point of view I don’t see at the moment a way how AI would actually work in isolation without human supervision and guidance.

So if we just, like, try to demystify all these myths to my mind, we are not at the point where AI could do everything itself, but if somebody not improving his personal level of expertise used to do like boring, repetitive kind of work, this kind of people in danger because most likely these people might be replaced because it’s not a secret that some developers don’t like to follow the trends, learn something new, become more expert in what they’re doing, and this might be a problem for them because AI may equally fast generate the code as they do with the same level of quality, and for them it would be a problem. If somebody is an expert, a good programmer, architect, artist, etc, say in the first instance consider these tools as time savers so they can do more in a shorter time.

Kara Goward: So it’s more about working with you and improving your skills than replacing your skills.

Andrei Salanoi: Yes, for example, just a few examples. I’m an artist, and I’m supposed to create an illustration just like a naive estimate may. It took me like few days to build an art. So I can actually present it to someone or I can generate an image using stable diffusion journey, whatever generic AI and then spend let’s say half a day on polishing this masterpiece and instead of like doing something for a few days, I’m getting a result with a half a day and I’m getting more or less the same money, even over the time we will start to get less money for this like piece of work. But by doing a lot of these pieces of work we still can actually earn a lot of money just because we became more productive and the same story with developer.

A lot of time the developer is just coding, they just type in the code here and for a minor fraction of their time, they are thinking about the solution what they need to do once they make the solution, build some solution in their mind they need a lot of time just to write this text which will be later converted in a programme. Now they can ask. Whatever, to clean its steps, just revisit the outcome because it’s done very quickly. The person doesn’t type with such speed and if they are happy, then just they commit their work. They did the same amount of work in the way they would do this manually but they are using AI tools.

Kara Goward: Yeah, there’s so much potential and I think if we work with AI, it can, it can grow a person, grow the skill set. So I suppose going back to kind of Godel and AI, I know we now have an AI community. We have a growing interest in what we’re going to do with AI, so I suppose, why did we start the AI community?

Andrei Salanoi: Well, there are a few reasons we observed like growing interest among our clients. So there should be an authority inside the company where you can actually come and like ask for assistance for advice, for expertise for whatever. So there were no places where you could come and ask your questions or ask for assistance. On the other hand, we go to such a I would say state of things in Godel that various initiatives started to appear in different divisions, initiative groups, etc. And nobody had a holistic vision of what was going on? So there should be a need to have once again some group of people who know what is going on.

As a company, we need to follow the trends. So somebody needs to be in charge of this and all these reasons first like the company to think about getting some structure in their organisation which will be responsible for it. And at the moment it’s called AI community. Maybe over time, I’ll be called another way, but this is like the group of enthusiastic group of managers who actually now act as a centre of expertise regarding this topic.

Kara Goward: And how many are in the community at the moment?

Andrei Salanoi: So if we talk about like board members and the guys who are deciding what to do, it’s about like 17. So it’s like VP, C level people heads of divisions, head of R&D function. Yeah. In terms of active participants taking part in the discussions, it’s allowed, it’s about like 70 people. And in terms of what we call it spectators, participants of public events, probably it’s measured in hundreds.

Kara Goward: And what is the goal of the air community and how do you hope that will evolve?

Andrei Salanoi: In the first instance, spreads the knowledge among the company. I would say encourage people to learn what is AI, try to find opportunities to use it in a project, maybe during daily routines and start thinking in another way how they would do things using AI tools because sooner or later that will be a normal daily activity, as we’re using cell phone now. Yeah, because like 30 years ago, there were no cell phones, so nobody was thinking about how life would change. Now we start to think about where we can make use of AI.

Kara Goward: Yeah going back to like with our phones, who would have thought, as you said 30 years ago?

Andrei Salanoi: Yeah, no phones, mobile Internet. So things were sorted out in a completely different way.

Kara Goward: Absolutely. I think yeah, it’s just completely progressed now. So what tools have you introduced so far internally that can help colleagues with questions related to AI assistants, for example?

Andrei Salanoi: This tool is the one which is built by Jorge Garcia de Bustos. So-called like working JorgeGPT. This is like the implementation of well-known pattern or search question-answering systems in AI. Now we’re actually trying to adopt all kinds of AI assistance. For developers like Copilot, and Jet Brains, AI systems continue. So this is another kind of big piece of work we need to do. Other than Jorge’s solution, there are a lot of ideas which have been implemented now, but they are not at such a stage, so we can actually put them on a table and present them to someone.

Kara Goward: And looking outside of Godal externally, what impact do you think AI will have on our projects and processes at Godel?

Andrei Salanoi: Well, I would say that if Godel does nothing, we will suffer very clearly from this mistake because as we see other companies doing similar businesses go down, they’re not, they are publicly announcing as they start adoption of various tools and the primary reason for them to do this is to become more productive. Maybe to some extent do more for a smaller price and if we want to stay in the business or increase our share in this market, we need to be on the frontline. So we need to be aware of what is going on, know how to use it, what value to get, how to sell this to clients, etc, how we can solve clients’ problems with these tools. So this is the main reason for Godel all to be so active and like keeping track of what is going on outside.

Kara Goward: Yeah. So I suppose going to the client side, how are our clients reacting to, you know, a change in technical landscape and how AI is fitting into that well?

Andrei Salanoi: I would say different reactions, different extents from like let’s do it and we will never go to do this. This is very dependent on the personality of decision-makers like CTOs and Heads of Engineering. If the person is open to all new initiatives, is ready to experiment and see. So this usually is a conversation to find a common ground and try to do something. If it’s a conservative person who would like to keep everything as it is now because it works, so don’t touch it. So it’s a very difficult conversation, it’s difficult to convince them. Also, the size of the organisation matters, because it’s like a small team, they can’t actually afford to experiment because these are not so big spending on licences in comparison to the organisation, which has hundreds and thousands of people. So usually they introduce these processes for wide groups of people and proper planning, getting funding, etc takes a lot of time.

I see at the moment clients are mostly interested in like well known tools like AI assistant because they see it as a way to improve productivity. But we’ve also started to get interested in building something. I mean from a business point of view, using AI technologies. But it’s like rather than in the experimental phase. What really happens now is the adoption of the tools which can actually make developers more productive.

Kara Goward: And what tools do you think could revolutionise coding?

Andrei Salanoi: I’m not going spoil the secret (laughs). Well-known co-pilot, jet brain, assistant open source alternative, which is called continue. So these tools more or less do the same. So they actually allow developers to be focused, not switch from The programme where they create the code ID to browser in order to Google something and find like exact match to that question.

So now they can actually ask this assistant all these kinds of questions and as soon as this assistance was trained on a large amount of data it most likely may have similar information already built in this, let’s call it knowledge base. So the relevant answer could be provided in the same application in the same context as the developer is working now and it’s actually it makes the life of developer easier because there is no context switch between different tasks and you can actually stay focused and productive.

Kara Goward: It’s speeding up processes isn’t it?

Andrei Salanoi: Yes, in my personal experience, if you are not an expert in certain technology, because recently I had the need to create like a JavaScript plugin for the browser because my area of expertise is .NET is not something you can actually use to build a browser plug in. The AI systems was very helpful for me because he was suggesting as solutions which would take more time for me to find an Internet and I just explaining the intention of what I’m going to what I would like to do and it was offering me a solution if the solution was not good enough, I was explaining what is wrong with this solution and another solution was offered and then like getting this piece of code I managed to like build fully working plugin which solved my problem.

Kara Goward: That’s interesting. Do you have to tell it when it needs to improve or change?

Andrei Salanoi: Yeah, the idea is that, like, you haven’t conversation in the chat form with this, let’s call it AI. So I would say, I would like to create a plugin for Firefox which is going to do blah blah blah, please just create a skeleton of this application. It creates your skeleton and instructions on how to install it. Firefox and you say, OK, so in addition to what you’ve done, I need to change this piece of code in order to work in this way and look like this. So you start to actually clarify the details of this solution. Maybe then you have another idea, so you ask something else. So this is like a collaborative process and probably from this point of view, it’s clear why it’s called copilot because it’s like impressions that somebody is sitting next to you and not mixing you. It’s not like some virtual AI system which like doing everything without interacting.

Kara Goward: Yeah, that makes sense. So I suppose my last question and I know we’ve made a lot of predictions at Godel, but while I’ve got you here, can you just give us and our listeners what’s your biggest prediction for AI technology this year?

Andrei Salanoi: It’s like it’s always a guess, it’s like a probability and I’m not going to be right, but what I’m sure it that all these tools and AI systems will be widely adopted very quickly and it might be like a normal part of the process as we use tools like. Brains ReSharper, which like almost every developer could not imagine his life without this plug in. So very quickly we will get to the point that developer will probably refuse, or at least complain that they could not work with without this kind of AI system, because they don’t feel their self to be productive. This is something we can actually expect will happen very quickly, maybe within this year because it’s evolving so quickly. So I will not be surprised if next year it will be like a mast as the default setup for any development environment where you consider to like to do something on a serious scale.

Kara Goward: It’ll be so interesting to see where we are this time next year.

Andrei Salanoi: So you can actually record this and we will compare where you are. We’ll do that.

Kara Goward: Part 2 coming next year! But thank you so much today for your Andrei.

Andrei Salanoi: Thank you. Bye.


Kara Goward: Huge thank you to all the listeners out there who’ve listened to this instalment of the Godel pods. If you like what you hear and would like to know when we’re releasing even more episodes, please just subscribe to the Godel page on Podbean or Spotify.