WILL AI REPLACE DATA ANALYSTS / DATA SCIENTISTS / DATA ENGINEERS?

one of the questions I get asked most by you guys  is how I think AI will impact the field of data   analytics and whether or not AI will replace data  analysts completely so to answer the AI related   questions once and for all at least for now let me  quickly share my thoughts on this topic with you hey my name is mochan and I work as a data and  analytics analyst within the financial services   industry so the big question will generative  AI replace data analysts I'd say no but it   will certainly redefine the data analyst role  since chat gpt's release in November 2022 I've   seen so many articles on how data analysts might  be replaced by AI based mostly on the ability of   these large language models or llms to write code  for example chatgpt recently released the plugin   called code interpreter which lets you upload data  files and invoke python to perform descriptive or   regression analysis and even create visualizations  and you can do all of this obviously without   writing a single line of code would this be  the end of the road for data analysts then   probably not I don't think so as impressive as  the capabilities of code interpreter or GitHub   co-pilot or in general llms like chat jpt or  bardar there are some significant limitations   at the moment the main one being security I  work at a big organization and I would never   get the approval to get any confidential  data within my organization outside the   company's firewall into a large language  model that the firm has zero control over   so the current other and probably most popular  alternative is for each company to build their   own large language model now I'm not an AI expert  but training and fine-tuning such models including   the expertise and the Personnel required around  it would incur massive costs so would really only   be a viable option for very few organizations  but say you work for a company that has its own   large language model if your job is solely made  up of repetitive business as usual tasks such as   creating Financial reports by pulling in data  using SQL queries and then producing reporting   in Excel using pivot tables and pivot charts then  yeah generative AI will very likely replace you   but if you're a data analyst who has domain  knowledge someone who focuses on business   outcomes someone who identifies and solves  business problems then you should look at   generative AI as your friend not your enemy  just think back to the turn of the Millennium   the early 2000s when a bunch of new business  intelligence tools were introduced with one   ultimate purpose to allow non-technical people  to leverage their domain knowledge by allowing   them to analyze and present data without the need  to write any code I don't know about you but this   sounds suspiciously similar to the purpose of  introducing generative AI to me if you are not   familiar with the roles and responsibilities  of a data analyst you might think that they   just sit around writing some reusable code and  produce reports and dashboards on a regular basis   but being a data analyst is much more than that in  my opinion the greatest value of data analytics is   the ability to answer ad hoc questions unforeseen  unpredicted business critical questions complex   multi-dimensional non-linear questions for  example why is the conversion rate for the new   savings account lower than the other ones why  did the customer churn rate triple last month   why did the sales for the best-selling product  decline suddenly these are not straightforward   questions that can be easily answered following  some decision tree to be able to answer these   you'll need domain knowledge say for example a  small business owner running a coffee shop wasn't   just magically create and develop an end-to-end  data Pipeline with automated and insightful   visualizations and dashboards as one of the end  products sure chat GPT or GitHub co-pilot can help   you write optimize and debug code but if you don't  actually know how to code how do you even ask the   right questions to prompt these AI tools how would  you even know what to look for now of course you   could just fine-tune the company's model with  more and more precise data but it'd be very   difficult to train the model on multi-dimensional  questions that were never actually asked before   so for these main reasons I just went through  and probably for numerous others I don't see   generative AI replacing data analysts completely  to me the key lesson here is to take advantage   of AI to help you become more effective and  productive whether you're thinking about becoming   a data analyst or you're already working as one I  strongly encourage you to learn how to use these   new shiny AI tools like chatpt GitHub co-pilot  or board instead of being in constant AI fear you   could just make yourself AI proof at the end of  the day AI will not take your job but people who   will utilize these AI tools effectively they will  I have no doubt that generative AI will reshape   the field of data analytics change is constant and  normal and the question to me around AI is less so   about will it replace data analysts and more so  about how can data analysts harness the power of   AI to improve their workflows in my opinion a  typical data analyst in the future is someone   who can do the critical and analytical thinking to  solve business critical problems combined with the   ability to Leverage The Power of generative AI  tools to improve their data analysis workflows   so that's all for now on AI and data analysis  if you like this video make sure to check out   some of my other videos right here thank you so so  much for watching and I'll see you in the next one [Music]

Channel: Mo Chen

Views: 10,873

Keywords:

Id: NC7Maf-DYXA

Length: 7min 38sec (458 seconds)

Published: Sun Jul 30 2023