Jamie Lee Coleman: Magic of Automation and Everyday Chores

In this talk, Jamie Lee Coman, a Developer Advocate at Sonatype, explores the transformative role of automation and artificial intelligence (AI) in software development and everyday tasks. He discusses the historical context of AI, including notable milestones such as Alan Turing's contributions and the evolution of technologies like IBM Watson and chatbots. Jamie highlights the current state of AI, examining its advantages and potential pitfalls, particularly its impact on job displacement and security concerns. He also shares practical applications of AI tools that enhance developer productivity, such as code generation and documentation automation. The session underscores the need for ongoing vigilance regarding cybersecurity, especially as AI is increasingly exploited for malicious purposes, while emphasizing the importance of human oversight in tech innovations.

[Music] well welcome everybody to the next talk this Nick talked here um about the magic of Automation and everyday chores and with that we're welcoming Jamie Lee Coman to this talk and he he's a developer Advocate at Sona type and he was at IBM previously and he's based in the UK and he focuses on discussing the importance of security in software enhancing developer productivity and raising awareness

about the energy consumption of technology and uh I also had had the pleasure of working with Jamie back at IBM so very happy to be back here hosting him today so everybody Welcome Jamie and the stage is all yours thank you Mary thank you so good morning and well afternoon I suppose if you're in liania um hello everybody as Mary mentioned my name is Jam Lee Coleman

um and today I'm going to be talking about the magic of Automation and how it can help us as developers um so normally yes I would have to um introduce myself Mary's done a very good job at that so I won't give you the whole Spiel but yes um I'm a developer Advocate at Sona type I used to work on mainframes at IBM when I first started

my career which was very interesting coming from University and then I worked on websphere runtimes and then I was a developer advocate for them for a short while as well so um this is the part where I'd normally ask you to put your hands up and tell me if you've heard of Sona type before obviously I can't see everybody so I'm just going to presume everybody knows

who sonot type is right um if you're in the Java space you'll probably use our stuff all the time because we run the biggest uh Java repository uh called Maven Central we have done for quite a long time um just something we do for the they develop a community from C type um if you're you might be familiar with some of our other products we were the

first ever artifact repository in the world so Nexus you may know that um there's nexus firewall life cycle all these things but I'm not really here to to talk to you about our products today I will talk a little bit about how we use Automation and AI in our firewall products but just just briefly just to show to tell you how we take advantage of this stuff

so what's on the agenda for today um first we're going to talk about the artificial intelligence Revolution it's not something that's happened over the last few years I guarantee you that um I'll talk about the current state of AI then what I did is I asked chap GPT to automate and generate some slides for me all I did did was change the word advantages to disadvantages in

the question so we'll see if there's any bias or anything there um then I'll talk about if we should be worried as engineers in general some practical uses for AI um how we use AI at sonot type and then just kind of my conclusion in my opinion on where AI is going and how it's going to affect us in the future you'll have to bear with me

because I've got a bit of a cold I've been traveling constantly to conferences um I was in Poland on Friday and everybody seemed to be ill so um if I if I sneeze and I um I cough I do apologize so the AI Revolution um it's a big thing now every conference is talking about AI everybody's talking about AI um but AI is not a new thing

not in the slightest um so let's rewind quite a few years back to 1950 um World War II has just come to an end and Mr Alan churing if you don't know who he is he cre he built the first ever computer in the world technically right um he didn't design it I think it was actually a Polish lady who designed it but back then to build

the to create the idea of a computer was um very expensive and really there wasn't enough usage for it but then obviously World War II came and basically the British government employed Alan shuring and a team of Engineers to build a computer to crack the Enigma code which was a code that the Germans and the AIS forces used to communicate right um so fast forward five years

after World War II um Alan churing already knew in 1950 that computers were going to get very very clever so he designed something called the churing test I'll explain a little bit about what that is later um logic theorist um this was basically a program that proved um 38 out of 52 mathematical theorems um this is kind of maybe the first kind of level of some kind

of intelligence we get from computers where it's able to you know complete mathematical problems that humans struggle to do right um fast forward we have the Eliza chatbot this was the first ever chatbot in history again a long time ago 1966 right then we fast forward to webots this was basically the first ever humanoid robot so a robot that was built in Japan um so again still

quite a long time ago we've got robots at this point in 1972 um fast forward to IBM deep blue I'll later but that's a very very clever U machine which can be kind of considered artificial intelligence and then we go to the romba now I have a Husky and we've just got a new puppy which is surprisingly not bothering me at the moment um because my wife's

gone out but um really the Rumba is well is amazing for me because it cleans up all the dog hair in the morning before I even wake up right um but this was the first bit of home technology that had AI inside okay okay um so 2002 we've already got AI in our homes okay then we move forward to IBM Watson this was really a pioneering kind

of AI um that IBM created back in 2011 still around today fast forward to Eugene goodsman um this was another AI um basic an AI system and I'll explain a little bit more about that and then of course what made AI popular today well I think AI personally and chat gbt is what's a AI popular with people all around the world for the last few years so

the churing test if you don't know what the churing test is it's um originally called the imitation game if you haven't seen the film The imitation game go and watch it it's an amazing amazing film um but basically it was a test to tests a machine's ability to exhibit intellectual behavior um equivalent or indistinguishable from a human Okay um so if a human could not distinguish the

difference between a human and AI then the AI essentially would have passed the test okay um like I mentioned IBM deep blue they started development on this in 1985 so again quite a while ago um basically they wanted to put it up against the best um chess player in the world at the time um Gary casb off and um basically IBM won two of the matches but

Gary won four right but then IBM being IBM they were not going to let um that machine lose so they upgraded it in 1996 seven once again and they challenge Gary now out of six matches over several days IBM actually won two of them the champion won one and there was three draws so at this point the machine is now getting to the point where it's better

at playing chess than a human okay and not just any human the best chess player in the world time so fast forward a little bit um we have IBM Watson this was a great system I used to use it at IBM um you could ask it stuff in natural language and it can comp computate stuff from that um it was initially developed to for a quiz show

I don't know if you've ever watched this but IBM put pitted Watson against the best quiz Masters in the world and it came first right and it won1 million US dollar I mean to be honest the electricity consumption to power the thing for that quiz show probably was more than that but they weren't doing it for that for million dollars right they were doing it to prove

that AI had now got to a point where we could train it to be better than some of the best quizz Masters in the world um and then in 2013 it was used in healthcare and I thought this was an amazing idea because for examp I think AI in healthcare is really amazing because at the end of the day doctors can't learn about all the new treatments

and new diseases and stuff that come out on a daily basis right no doctor in the world will know everything um but we can feed this kind of information to systems like AI enabl systems and they can then help the doctor basically diagnose patients and also figure out out the best treatment plan right so this was really really really cool in my opinion um yes so Eugene

goodsman um possibly the most advaned chatbot of its time it was developed in St Petersburg in 2001 um and I find it funny now because they basically based this on a a 13 Ukrainian boy um which is just silly but the reason they did that is so um if it said anything silly or stupid then you could forgive it right because it's just a little child at

the end of the day um this this uh chat bot was very very Advanced so in 2005 and 2008 it finished second insuring test competitions right so insuring test still a big thing and in 2014 on the 60th anniversary of Alan cherin's death it convinced 33% of the judges that it was human right and some people do declare this as passing the CH test okay um so

let's fast forward a little bit to the current state of artificial intelligence in today's world um we've probably all seen the videos online of um deep fakes all these celebrities doing silly things like there's even an Instagram uh account just dedicated solely to uh fakes of Tom Cruz and they are really good like some of them are really indistinguishable right so we're all aware of those kind

of things but have we actually tried using them and my first step into this world was using Adobe genri fil amazing I remember when they came out of this about a year and a half ago qu it can it can really touch up any pictures it can add anything in it's a bit scary because we don't know what to believe is true anymore all these Instagram people

going around the world saying look at me I'm here when they might me in their bedrooms you don't know these days um but this is a really really amazing toour um but what really has made AI popular today is chat GPT right my grandma and my mom can use this anyone who can type sentences onto a keyboard and click to get to a website can use AI

right whereas my previous experience with AI IBM was you know you have to you have to connect to apis you have to set up a lot of different things blah blah blah blah not something a non-technical person could do right whereas now anybody can use AI it's really really quite impressive so open AI chat gbt um announced in 20 22 it is it became the fastest growing

consumer application in history um there's lots of chat Bots based off it I know I've got Google bard there now I think they called it Gemini now or whatever they've got a new version but um yeah there's lots of different people implementing AI these days and some of it based on what open AI have actually done um but it's not all happy days right so in order

to train this stuff at the beginning um open a I outsourced a lot of its training to Kenyan workers for example and it was paying them less than $2 an hour and why that might not be a really terrible wage in their country if you think what they would have to be looking at because there's a lot of evil humans around that are put disgusting things into

these chat Bots they had to go through all of that stuff and make sure that the model was trained right and they basically describe the experience as complete torture now I don't actually know what they do nowadays I presume they W do that anymore um but again so all these Technologies we use all this AI we're using all this automation someone's had to actually create that create

the automation for you create the AI for you so um yeah just give a thought to the people that have to do that forward right so uses for AI and developers right so um we do with this report so you can check it out um you can see the link at the bottom just just Google for state of the software Supply but what we do is we

talk to a lot of our customers we talk to different people in the industry and try and create a report on how the world's changing how open source is changing right and for example AI came up a lot this year um if I take the first point from the the last section Junior developers benefit from having AI machine learning tools to answer quick questions and provide insight

into technical terms and jogon I think that's an amazing use for AI personally I did a talk in Germany this week um around static analysis tools and how they can help Junior developers be trained quickly um to get up to a really high standard okay and I think it's the same for AI and I can see a lot of AI eventually moving into these kind of static

analysis tools um but there's lots of different things we can use it for now these are some of my favorite things um that I like to use ai4 as a developer and code Snippets is a really good one works with Java python C++ blah blah blah blah um and it basically creates error prone and performance optimized codes it's really really good um so do check that one

out that's amazing we all know of co-pilots I don't really need to um give you an overview of that but again it integrates to a lot of your IDE so if you're you're allowed to use it in your organization I definitely recommend because it I find these things really really good if you're starting on a new technology or you're starting on a new project and they can

help you get the Bare Bones or get started really really quickly right I hate Rex I don't know how many people here have to use it I hate it um so Auto reject is something really good um you can basically convert English into Rex using natural language processing really really good tool if you hate Rex like me um it takes me so long to do the smallest

amount of thing because I don't do it very often right and I think that's the point a lot of these developer productivity tools are really really good for tasks that we're not very good at stuff we don't do very often I think if we're very good at a task then it's probably going to be a while before the AI tools are really really going to be able

to help us there right and and documentation something we all have to do all the time um and it's just a pain like we want to focus generally on Innovation we don't want to have to be writing documentation all the time so minifi um is very good at generating documentation and it can understand quite complicated functions um and generate documentation from them so do do check that

out um there's a few there's another thing um I won't give you the name of it because I thought it was open source until it kicked me off and told me it was a trial but something I've been using AI for which has been amazing was um so recording uh taking down notes in Zoom meetings and things like that I used this AI tool um it was

amazing it would even not take down notes for stuff that was not related right so if we started talking about someone got a new car in in the team or this it would just emit all those notes um it would structure it perfectly to the point I was pretending for probably 3 weeks that I was just the most amazing Note Taker in the world everyone was like

wow that's amazing and then my trial expired I was like no um so yeah tools like AI tools was things like that are really really amazing um one thing so another AI tool and here's an example um that we've been using it's not this exact example but this is a way you can do that for free using open source um is basically creating an internal chat bot

right so we get a lot of questions from sales marketing different parts of the organization people that just work in different technical teams and asking about how this works how that works blah blah blah so what we did is we bought an off the-shelf AI thing that we could just train with our own data and that saves us as Engineers a lot of time because that is

now the first point of call that everybody goes to when before asking a question whereas before we used to get bothered quite a lot salespeople and things like that so one way you could do that yourself again using open source stuff um if you're in the Java space and don't be afraid of AI and Java by the way I am a Java person um I know this

conference isn't a Java conference specifically but don't be afraid of AI and Java Javas are actually quite performant with AI um I've seen quite a few other people do benchmarking tests against Python and it's beat it in quite circumstances um I'll copy this link over um but again this is all open source so you're using Lang chain for J micro profile which is a microservices framework for

Java Jakarta e which is what Java e used to be um open Liberty is a runtime that IBM develop again open source and hugging face if you don't know what that is it's basically an online kind of repository for storing um models and things like that and large models for AI systems and but what this does is again you can just go and train your own model

you can basically take this off the shelf example um deploy it uh point it at that specific model on hugging face and then have people in your organization ask this chatbot this chatbot questions rather than having to keep coming to you and things like that um they've created the people at open Li have created a really I'm I'm a little bit biased because I used to work

there um but they've created this really great article which explains how it works what Lang for chain what chain for J is how to get started and again it's a very simple web application all you do is type your messaging click Send boom um but then you can go and train it on your own model on your own information um so yeah it could be really really

really useful for you all okay let's go back to the presentation oh let me copy that in um if I put it in the private chat I don't know if this is going to go to you all but if it does or anyone marry or whoever and if you could send that to the audience that would be amazing just so they can have a look at it

and try and create their own kind of automated way of answering questions right so as Engineers can focus on what's important long have I got uh I've been live for 19 minutes okay plenty of time so should we be worried as Engineers or first of all should I be worried I'm a developer Advocate my job is to do this all the time um generally I don't get

chat GPT to generate my slides I may ask chat gbt some questions which might be useful but I don't get it to generate presentations until now right so first question I asked chat gbt was give me the advantages of artificial intelligence give me the advant vantages of AI um in presentation slides so the next few slides are generated and then I ask the same question but change

the word advantages to disadvantages um and then generated those slides so I just wanted to see if anyone here can notice any bias with these right I'm not going to read all the points in these slides but I'll read the first point um and I agree with most of the things in this presentation so first of all increased efficiency right AI technology can automate repetitive and Mund

tasks freeing up human resources for more complex and creative work yeah I'd agree with that um enhanced decision making AI algorithms can analyze vast amounts of data quickly and accurately enabling informed decision- making yes so like I mention with wats and health things like that like really really useful for those situations and improve customer experience AI powered chat Bots and virtual assistance provide um instant and personalized

customer support enhancing the overall experience yes I can agree with that I've used some AI systems before when I've been on the phone and they know it's me instantly and get me to the right place so really good um cost saving so AI can help businesses reduce costs by automating processes eliminating the need for manual labor I agree with that and again that might be one of

the points when some Engineers start to get worried because AI can automate lots of undan tasks personally I think that's a good thing but we'll talk about that later um Innovation and creativity oh by the way feel free to ask me any questions while I'm going along I'll I'm more of a free flowing speaker so I'm happy to just try and answer the questions as I go

or we can save them to the end um but yes Innovation and creativity AI enables re researchers and scientists to tackle complex Problems by leveraging competitional power and advanced algorithms similar to one of the other points before and one the other side but yes I agree with that as well um improve safety AI can be employed in various Industries to enhance safety and reduce risk yes 100%

um we've seen that with self-driving cars they're generally a lot safer than humans I think most of the risk on the road is generally humans and most of the problems self-driving cars have is with humans not the actual self-driving car itself um but I am also a petrol head so the day that we're not no longer allowed to drive vehicles will make me cry but um I

probably won't be around for that and personalized experiences again very similar to one of the points we had earlier and AI can analyze user data and preferences to deliver personalized experience is such as customized recommendations and content very similar to one of the previous slides but again I agree with that so the conclusion of the advantages of AI artificial intelligence offers numerous advantages across lots of different

industries from increased efficiency enhance decision making to improve customer experience cost savings um AI is transforming the way we work and live and embracing AI Technologies can unlock new opportunities and drive Innovation okay cool completely agree with that now um see if anyone can see the bias between the two again all I did was change the word advantages to disadvantages okay so job displacement Ai and automation

technologies have the potential to replace human workers in various Industries yep completely agree with that I don't think we need to be worried I'll explain Ry in a bit um lack of human judgment and intuition AI systems lack the ability to exhibit human level judgment intuition and creativity yes um the creativity yes I suppose I'm not that creative um it's more about how what your creativity and

how you want to get it to do it but yes I I agree with that privacy we've seen this before when chat GPT and other things first came out um AI systems often require fast amounts of data and where they get that data from can be a little you know wishy-washy um so that's not always good dependence on technology overreliance on AI technology can make societies vulnerable

to system failures I think that's just not a specific thing to AI personally I think that's just an overreliance on technology as a whole um I don't think AI is any more susceptible to failures than overreliance on AI than just general technology personally um ethical config considerations AI can algorithms can perpetuate and amplify existing biases which is what I'm trying to show you here and lack of

emotional intelligence AI systems cannot understand or emphasize of human emotions and social cues right um while that is true um you do have sentiment analysis in text and speech detects and things like that so slowly getting to the point where it might be able to understand if someone's sad or angry um but yeah I understand uh unemployment and income equality as AI automate certain jobs it can

ex exaggerate ex exag speak English income equality and create a divide between those who benefit from Ai and those who do not that I 100% agree with and I'll explain a little bit more about that later and lack of accountability who's responsible if an AI makes a decision and stuff goes wrong right if we're got a self-driving car um a child runs in front of the self-driving

car and the self-driving car is AI enabled it has to make the decision to Swerve left or in case of two-thirds of the world swerve right onto the opposite side of the road um and another car might hit it so that that could be a collision people could get hurt or die there or it could go onto the pavement sidewalk and wherever you're from and there could

be people on there so who is accountable for the decision that that AI makes in that circumstance right so conclusion of disadvantages while AI offers numerous benefits it is important to be aware of its potential disadvantages and addressing these challenges requires careful consideration regulation and ethical Frameworks now um I'm not sure sure any can anyone there see the difference between the two points right so and I

gave you a little hint there of what the difference was so advantages this was disadvantages okay so the clue was in how many points as you can see when I asked it disadvantages it only gave me two points for each slide and it could have easily found a third point right I found third points for all those arguments where as the AI itself it fought oh I

don't want to talk too bad about myself right so it only gave me two points and you can already see there is some bias there okay um should I be worried maybe I should be a little bit worried it was quite good at making those slides I think it's going to be quite a long time before a robot is going to inperson conferences and standing on stage

replacing me um could happen in my lifetime but um so I maybe I should be a little bit worried and should you should General developers software engineers in general should they be worried about AI um my personal opinion is no okay so yes for a totally different reason um in regards to security but no in terms of like ad jobs okay I think like I mentioned we're

a long long way away before this this kind of robot is able to do a lot of tasks around the world um I quite like this quote um to replace programmers with AI project managers need to accurately describe what they want so I think we're safe for quite a while right um but the reason I say we shouldn't be worried is um our job is essentially to

automate the world okay that is what software Engineers do we take tasks that are manual and we try and create an automated process to make life easier for the person using our software right so we our job is essentially to a aate every other person out of a job that is what we do okay now unfortunately that will mean we will be the last people on this

planet still working okay because someone's got to keep all the it systems running someone's got to create the new it systems um so we're going to be the last people working on this planet but I don't think we should be worried personally because first of all there's lots of other automated jobs that AI are going to do W before automating our jobs first right there's lots of

they can do but what I would say is someone an engineer that knows how to take advantage of AI might take your job right the AI system is not going to take your job as a whole someone's still going to have to run that AI system generally right and if you're doing a job where an AI system can easily take over it then you probably should be

looking for a new job and Tra retraining yourself anyway okay um if your jobs are easily replaceable then you you should be trying to retrain yourself anyway but I honestly think if you are you can take advantage of all the because that's what we should be doing anyway taking advantage of all the new technology that comes out so if we can take advantage of it and make

ourself better Engineers I don't think we'll have to worry about losing our job in the slightest but like I said if you're doing a very mundane automatable task then you should be trying to find a new job anyway because yes s Engineers will eventually automate that task right and that's a good thing because it allows humans to focus on innovation allows them to be creative rather than

doing repetitive tasks over and over again so the part we should be worried about as Engineers um the way the bad ways AI is being used right so cyber attacks have risen dramatically um over the last well since the internet was born right um 1992 and um nation states essentially preparing for the next War but it's not going to be physical it's a lot more to do

with hacking okay and the days are gone where it's a load of people in a basement trying to earn money yes that is still the case but a lot of these hacking groups are now nation sponsored right and they're trying to take down our infrastructure they're trying to take down our medicine production uh uh Logistics in terms of food deliveries things like that and they're not just

doing it straight away they will come in and wait in your system they've exploited your system and they'll wait there for the opportune moment to cause them most amount of damage okay so the days are changing U things are changing and we need to be a lot more prepared for this stuff um like I said cotype runs Maven Central the biggest Java repository um there's a lot

of things that decisions that were made at the beginning that help us reduce stuff so you have to own the domain for example if I'm putting up something from Apache I'd have to own the domain or. Apache to do that okay um there's different ways people are trying to get socially engineer into our system now because our firewalls are pretty robust as long as nobody makes any

mistakes and allows these people in by you know plugging into malware and a USB stick which we should all be aware is not a good thing to do um they'll try different things like dependency confusion where they'll just add a really long version number so you go and download the latest and all of a sudden you've got malware vulnerability they might do typo squatting so instead of

log for J it might be log j4 things like that right with Maven Central the first two are defeated because you have to own the domain but nothing stopping someone from creating a domain their own new domain paying for it and then registering that right so we do have Protections in place but again um it's still something that can happen whereas if we compare the node ecosystems

and the python ecosystems their repositories are full of malware we do take down requests on a daily basis to these repositories to get this stuff removed we identify stuff on a daily basis um quite a lot of things on a daily basis I mean like I mentioned Java is pretty safe um I think 99 or 97% of all the malware we find is through these two ecosystems

right so quite scary but stuff's hard okay I mean we didn't have to worry about this 25 years ago we were writing everything ourself whereas now we do okay we have to worry about that package we're downloading does it contain vulnerabilities how do we figure that out how do we stop the bad code being published etc etc so there's lots of different things we have to consider

that we didn't have to do previously and for Maven Central again no I think malware is only been in there once and I think that might have been a test by xenotype but everything's scanned with our commercial scanners so we automate the whole process of when you try and push it up we do a a load of automative process to check that there is nothing bad in

there and just to clarify the difference between a vulnerability and malware is malware is made intentionally to be malicious it's meant to be bad okay a vulnerability is generally a coding mistake that opens up a security problem generally a vulnerability but it's not malicious okay so for Maven Central um again if you're not in a Java space don't worry too much about this but what I'm trying

to emphasize here is we should make better decisions earlier on when building our applications choosing what libraries we're going to use rather than having to deal with the bad things later there we go so unfortunately we as developers are quite slow to change um if you're in the Java space you'll be familiar with the log for J vulnerability this was probably the worst vulnerability in Java's history

uh this was a few months after it came out of course we were Maven Central so we can see how many people are downloading that vulnerable version um and at that point a few months later we were at 51 million downloads right we fast forward to a few weeks ago um it's still being downloaded at a crazy rate 33% of all downloads of this Library it's a

logging Library that's embedded into lots of things um are still that vulnerable version 354 million downloads of Java's worst ever vulnerable Library still it's quite crazy right um that's not going to stop us using open source like I mentioned this this is a new thing that's happening because we are consuming lots of Open Source and in fact we did a study that said 90% of the code

we ship is not our code right it's code we have taken from dependencies that's a lot of stuff that we are not paying attention to and the bad people using Ai and things like that are okay so we take a average Java project for example 150 dependencies each one of those has 10 releases a year that's 1,500 updates to consider that 25 years ago we were writing

all that code ourself so we didn't have to worry about it too much um and we've all probably seen this this image but yes it generally turns around that a lot of these projects are maintained by a small group of people um and they can be very easily influenced um with AI and lots of other things to essentially turn not them personally turn malicious but convince people

to hand over projects and make them malicious essentially um and it's a lot so this is um spring web framework for example and as you can see we pull in one project here that pulls in all these and these dependencies and these dependencies and these dependencies so there's a lot of different attack vectors for the bad people to utilize right and the modern face of cyber crime

is it's difficult tasks detailed plots high technology elaborate schemes but in reality they go for Simplicity that this is why AI is their friends because they can use AI to very very easily get started doing malicious stuff so one of the first things I did with chat GPT and this is when it was just released is asked it to generate new malware in Java in Python in

node it did it in every language and I kept asking it to tweak it and it was tweaking it in different ways so again people are taking advantage of Automation and AI to attack us and do bad things um this is uh malware as a service you can go to this service and it will customize and generate malware for you which then goes into these repositories that

we use as Engineers Okay so we've got to be very very careful because the bad people in the world are taking advantage of the this AI technology and they're using it for very very bad things now why does that matter for us as software Engineers right well some crazy fracks here I've thrown these size around a few times but um in 2016 cyber crime um from one

study stated that it's a past the drug trade okay that was 450 billion US dollar a $14,000 a second that is equivalent to 50 of the world biggest nuclear powered aircraft carriers right so um crazy crazy am I still here can you still hear me Mary hope so um yeah so lot of lot of like lot of aircraft carriers right um that's got much worse 2022 what

we talking about Six Trillion that's equal to 620 of the world's biggest aircraft carriers okay crazy amount um if that was the GDP of a country it'd be the third largest GDP in the world okay and I've recently um over the last few months seen some more studies that say we're at about n trillion US Dollars okay so crazy crazy crazy amount of um money and what

does that get you you ask well it'll give you a year salary for 162 million teachers in the United States it will give you 36,000 celebrity divorce settlement that doesn't seem that much Co they must be getting a lot in these divorce settlements um lots of Toyota Priuses and of course what matters to us most 11.9 s trillion chocolate bars they're stealing all our chocolate everybody come

on we need we need to stop these bad people seeing have chocolate and if this guy was around today I don't think he'd be in the drug trade I think all that money around him would be a load of hackers because as we see with um drug dealers and Drug criminals they generally get caught put in prison and get killed um cyber criminals don't often get caught

they generally don't get killed um so and they make a lot of money so I think the world is changing right and like I mentioned AI is making this stuff a lot lot harder um back if we went to 2006 um this is some research from IBM um it basically said that we had an average 45 days from when some malware or vulnerability was announced before we

got attacked whereas what we saw with log for that was about a day okay so it's a crazy short window now we have to be prepared when people start attacking us and open source code is in everything it's in insulin pumps it's in aircraft it's in self-driving cars it's in trains okay so we're being we're we need to be responsible for the stuff we're shipping now son

type how have we used AI well we used AI before it was cool not not so far back as Alan churing creating the churing test like we haven't been around that long um but we've been using it for quite a while okay um I showing you this but basically this is kind of the cotype platform like I mentioned you'll probably be familiar with um reposit C type

repository uh life cycle is a thing that continuously scans all your stuff right your software development life cycle to make sure there's no bad stuff in there and but the one I want to highlight is firewall so firewall for example is like the name suggest it's a firewall but not a firewall to stop in any incoming connection it's a firewall that basically goes between your um public

repository and your internal repository um what we use here and we've been doing this for about five years I think now six years is we have an AI machine learning um system that constantly scans everything that comes into firewall um it observes every new code commit things like that and basically we have a like a security research team of 50 people so my advice with AI is

AI is great it's a great tool to make us better at what we do but don't replace humans with AI generally because then there can be quite bad repercussions okay um but this system itself has found two dates over 350,000 new malicious packages in the world right like I said 94.6 all of these are from node and Python and it's quite scary when python is the main

language people are using um to create AI systems nowadays so yes we need to be very very careful but essentially yeah firewall sits in between um it kind of goes if we've seen it before oh yeah but yeah if we've seen it before then um it'll go straight into your system if not the AI algorithm will look at it the humans will look at it if it's

bad or good we'll retrain the AI system constantly because the bad people and the ones that are using AI to generate all this stuff they're being very very creative of how they try and attack our systems okay to summarize and hopefully I can give you a few minutes for questions um this is my conclusion by the way um so if you have any problems with this you

know feel free to shout at me um AI is not going to take most of our jobs in software engineering anytime soon if you have a very mundane repetitive job try and retrain yourself anyway because it that doesn't sound fun um current AI technology has been around for a long time it's not a new thing okay so yes it's an amazing thing and it's come a long

way but it's not new in any way shape or form um it's become popular just because my grandma can use it right um although it should take a while to type something but she can type um AI can make us look much much more produ productive in our daily jobs but AI can be used for bad things and with any amazing technology you'll always have bad people

that take advantage of it right so try and find as much technology you can to combat the bad things people are doing with AI and generally a lot of the time that seems to be AI okay um and again try and utilize some AI systems in your in your current business like I mentioned the whole Chat thing we have for training on all our documentation our technical

knowledge so sales and marketing can just ask that it's it's amazing it saved us loads of time um there's some useful links here the history of AI history of software supply chain attacks things like that and we have a dashboard for log for J if you want to check out any of those things just take a screenshot of that and take any of those down and get

in touch some types around on all the different platforms some more cool stuff to check out there and if you're in the Java space check out the new Maven Central um but yeah there's some articles me and a colleague created that and if you want to find my slides um scan this code you'll get to my terrible website don't worry I'm not a front end developer well

do worry whatever um it's not a I'm not a front end developer my wife actually is a full stack developer and she looked at it and was like what is this so I'm going to like yeah write help her write a blog and she's going to fix my website for me but you can find all my slides and everything um if you scan that QR code or

just go to the link at the bottom and it's just a GitHub page I set up and with that I will end my recording which gives us about two minutes for questions so thank you everybody thank you Jamie Helly very uh you know full of lots of great information so yeah I guess at this point I do not see any questions coming from the audience um but

please you know we still have two minutes and if anybody wants to ask a quick question to Jamie see go out start I'm going to go outside for a few minutes cuz it's a rare Sunny Day in the UK yes I know in in yeah near you you're near London right yeah about an hour away yeah not too far yeah great yeah great so yeah I don't

don't see anything and it's been kind of quiet maybe it's just me maybe I'm just scary I don't know maybe I scared everyone too much not really it's funny too here in Chicago we are just starting the day and here in Europe too it's like in yeah I keep forgetting you're in Chicago in the evening now although you're probably still on European times right because you just

came from Germany so exactly I've been getting up like waking up at two and three in the morning so I figur okay no big deal I can get up I'm keeping my European time schedule for now so nice y yeah wonderful yeah let's see oh there's a question wait in the Q oh okay really quickly if AI is used for breaching security can we use AI to

protect us then yeah so kind of similar to what I said I mean IBM uses AI with their ex their Cloud providings um a lot to try and combat AI so what IBM is like sponsors Wimbledon big tennis com uh big tennis thing in London and they use AI there to kind of combat any incoming threats but like I said we use AI at zot type with

firewall to combat incoming threats as well so it's I mean to rely solely on humans is getting difficult but like I mentioned um I would never take a human out of the equation completely because that can lead to quite bad things and we don't want overreliance on computers because when they do go wrong then we're all a bit screwed right so yeah hopefully that answers the question

great I believe so yeah let me see if there's any response it looks like folks are happy um and I guess we are just at time so this is perfect if folks wants to reach out to Jamie uh Jamie has all the you know contact information uh from this site too so yeah so very good yeah well thank you Mary you've been an amazing host pleasure to

see you again it's great to see you too thank you thanks for a great talk thank you Jamie all right take care everybody thanks by everybody by yeah e