Get into Data Analytics NOW!

[xXOGLEGENDXx]

This YouTuber has valuable information and 100s of videos.

Alex The Analyst

 

[Regular Developer]

Most people have a misconception about AI and what it can do.

There are two main types of AI. RAG vs. Fine-Tuning. Most of what we associate with AI is Fine-Tuning. You ask and feed ChatGPT information to do something and it generates something based on what you feed it. In order for that to be effective you have to know specifically what to feed it and what to ask. When dealing with certain specific data sets, you still have to understand what the data is, how it has been manipulated and how it needs to be manipulated and interpreted. AI cannot do that on its own. ChatGPT is not going to replace us.

What would theoretically replace people is RAG AI. This is where a company's data is stored in a warehouse and a LLM (Large Language Model) is used to generate responses on its own and interact with us using all the data it has access to. Imagine your personality is downloaded on a drive and that drive gets plugged into a robot. The robot is programmed to read the drive and act as a vehicle for your personality. So instead of telling the robot (AI) what kind of joke to make, it already knows what kind of joke to make because it has access to your personality (data).

RAG is incredibly expensive and requires a lot of resources and power to be effective. Also it is reliant on the data it has access to. If it uses bad data, then the generations will be bad. That requires people to ensure the data has been properly managed and interpreted. AI cannot do that.

Will AI improve and be able to do these things at some point? Yes. But we are nowhere near the point where humans start getting "replaced". There is still a need for human data analysis, engineers, managers etc.

Also to add to the bolded, in order to train it, someone still needs to prepare the data correctly, which means cleansing it, getting it into a somewhat organized format, and then choose what the traits (i forgot what the technical term is for this) are that the model needs to consider when evaluating what data points hold the most weight for certain outcomes.

 

[xXOGLEGENDXx]

Also to add to the bolded, in order to train it, someone still needs to prepare the data correctly, which means cleansing it, getting it into a somewhat organized format, and then choose what the traits (i forgot what the technical term is for this) are that the model needs to consider when evaluating what data points hold the most weight for certain outcomes.

Exactly. Great point.
I believe the term is Feature Selection/Extraction

 

[Nokids]

Been a Data analyst for years now. Great field like a poster said you could be technical as hell but communication and telling stories (being an xray) will get you $$$. Here testing the market and here’s some salaries I been seeing.

Even with the TikTok fukkry it’s good to see what a nikka value is.. currently working for a French company paying $$$

 

[MikelArteta]

I'm a data scientist been working remote the past five years

 

[papa pimp]

Interviewed for a Data Analyst role at my previous company. But withdrew after finding out I would need to take a paycut from cyber. Screw that.

I'll stay in cyber until the wheels fall off.

even with all the bullshyt (burnout, nearshoring, scapegoats for any/every compromise), us folks already entrenched in cyber are blessed with the barrier to entry rising.

 

[Neuromancer]

Subs

 

[Cloutius Maximus]

That's pretty much mandatory even if you just maintain a simple Google Sheets page for any online side job


Google's course on Udacity helped get me started with excel and SQL after not needing it since high school

Otherwise just look up anything + free course and you'll find something for your learning style

There's soooo much free stuff on DA

edit

 

[#1 pick]

I’ve been in analytics on and off for 20 years. I’ve lead quite a few analytics teams. What’s critically missing from analytics is not the skillset but rather connecting analytics to usable insights and optimization. There are lots of people with amazing data science skill sets, but can’t tell a story for the life of them. They either lack business context or lack the skillset of insights and storytelling. I’ve met so many talented people with years of all types of data experience from analytics to infrastructure, python to sql to r to excel and all sorts of dash boarding, but I’ve only met a few people that have those skill sets and can also impact the business with insights and story.

Breh, what does this mean in layman's?

 

[#1 pick]

Interviewed for a Data Analyst role at my previous company. But withdrew after finding out I would need to take a paycut from cyber. Screw that.

I'll stay in cyber until the wheels fall off.

I feel like the wheels are falling off in my area of expertise

 

[O.T.I.S.]

I can’t do it justice in a post because it’s something that is learned over years but generally you’ll want to understand the client challenge and goals as it relates to their business so you have big picture understanding of what outcomes they want to achieve. Then you want to look at what you’re doing or what your team is doing to positively impact those outcomes. From there it’s a matter of what’s the story you want to tell where it shows you understand the client business needs, what’s working, what’s not working, start with big statements and peel the onion with granular data. Keep everything to a minimum, maybe 3 main points. There’s a lot of styles and ways. I can expand on it another day.

Actually sounds similar to things I do

 

[RamsayBolton]

brighter future than cyber security?

 

[Rev Leon Lonnie Love]

All that shyt is boring to me. I have a Masters in a STEM degree and worked as a Data Scientist for a while before I switched to Software Engineering. I think SWE is a better career overall.

Also, I think this AI buzz will die down within 5 years and no SWE jobs are really under threat of AI

 

[Gloxina]

This is true

 

[Gloxina]

Breh, what does this mean in layman's?

Essentially folks need to have EQ; need to know how to not only do the work and create reports, but communicate and be personable and know how to explain their findings and how those findings will effect the company. Make it more than numbers and graphs on a page.

Like back in the day when ppl would crack jokes about “the IT guy”. Great at his job and can fix anything, but can’t hold a conversation and has as much personality as a piece of cardboard.

What would happen if the IT guy (in this case data analyst or engineer) was as engaging as the guy in sales?