What is the Hardest Part about Working in AI with @samzee_codes

hi I'm Samantha Zambezi I am a data science technical lead for a large multinational marketing and insights company I am based in Amsterdam and I've got over seven years of experience working in data today I'll be answering the question what is the hardest part about working in Ai and to No Surprise its implementation and maintenance those are two things but I really believe that you cannot separate them there's various reasons as to why it's hard but for me I believe it's hard mainly because the formal education system did not adapt fast enough to changing industry needs back in 2012 when machine learning was like a big hype a lot of Institutions started offering this as a course or as like a whole degree program and in this degree programs a large emphasis was more on like the modeling side of it like understanding the theory behind the different machine learning algorithms as well as a programming side of it there was no follow-up as to what happens after you have a model that is giving you good results as well as what happens as data changes how do you then create an end-to-end system that's going to be able to maintain and manage this efficiently many organizations wanted these individuals that went through formal education to come into the organization and Implement these machine learning algorithms however the infrastructure was not there I was one of those that went into an organization and quickly realized that it is more than just modeling there's a whole infrastructure behind it infrastructure when it comes to data storage the deployment maintenance all of that even project management like it's a big system of its own and organizations are not static data changes processes change your machine learning models and algorithms need to be able to adapt to that change if you don't have efficient systems and technologies that are supporting these machine learning pipelines you start to experience large overhead data scientists start to really become like professional troubleshooters or they start doing more of the infrastructure work like data engineering ML devops and really stop focusing on the modeling side that is the site that's supposed to produce change however there is hope as now I'm seeing the AI Community coming together to talk more about these challenges about machine learning in production and how we can overcome them so there is now some Frameworks that are being developed some best practices that are existing a lot of companies have realized that they really need to start from simple and then move onwards to more complex systems unlike before it was like let's do like the most complex most fancy thing and we'll just figure out and see how it goes that did not work very well and that kind of demotivated a lot of organizations like if AI is actually bring in value so starting from small and building up on this as well as like knowledge sharing is really changing the landscape and we're gonna be seeing more on AI in production and more companies using and adopting machine learning so yeah that is the good news I want to recommend this book that I read about um three years ago and it is called applied artificial intelligence it's a handbook for Business Leaders it really speaks to the reasons why your organization might not be AI ready and things that you should be doing in order to become AI ready so I learned a lot from this book and I was able to take it back to my own organization and kind of identify like what things we should do better in order to get where we need to get to go check it out very great read and that is it for me foreign