that annoys me is the countless people online, in person, and even in my comments section saying “how AI will replace data scientists.”
I find this frustrating since it often comes from individuals who aren’t working in the sector, and it discourages those that could be great data scientists from pursuing this profession path.
Not to say, I firmly disagree with this view and consider AI won’t replace data scientists, at the least definitely not inside the following decade.
And that is coming from someone who has worked on this field for five years across a spread of corporations, and has seen what the industry was like pre- and post-AI.
I actually have zero concern about AI taking my job because it stands, and in this text, I would like to clarify exactly why I feel that and put an end to all this scaremongering.
You Need To Learn AI
Before we get into the actual “meat” of the article, let me start off by saying that I’m not a whole AI hater.
I exploit AI day by day, and consistently up-skill myself in AI because it is a crazy productivity tool for:
This technology is here to remain, and it’s essential to learn to make use of it; otherwise, you shall be left behind.
Competency with AI tools will turn out to be the norm, just as everyone is predicted to make use of email nowadays or know Microsoft Word.
AI won’t replace data scientists, but a person with fewer technical skills but greater AI proficiency likely will.
As an information scientist, it’s essential to be well-versed in tools like:
And so many more.
These will turn out to be staples in our industry, similar to Python has turn out to be the lingua franca of machine learning.
It’s inevitable, and it’s essential to get on board the ship as soon as you possibly can.
There Will Be Larger Problems
Let’s break down the talents AI might want to develop for it to completely replace data scientists:
If AI mastered all these skills to a level higher than a current data scientist, what job wouldn’t be gone?
If this happened, now we have far greater problems to fret about, almost singularity-level problems, and your concern about whether you need to go for an information science job will pale as compared.
If data scientists are replaced, there’ll likely be greater fish to fry in our lives than simply worrying about our careers.
Lack Of Mathematical Reasoning
One thing AI greatly lacks is mathematical reasoning.
I’m not talking concerning the layperson maths that the majority people ask AI like:
- .
What I mean by “mathematical reasoning” is the power to resolve unsolved mathematical problems.
For instance, AI currently can’t solve the Riemann Hypothesis since it lacks the creativity and conceptual reasoning to make a serious breakthrough in pure mathematics.
The Riemann Hypothesis is an extreme example because it’s arguably the toughest problem in existence in the mean time.
Nevertheless, it shows that AI hasn’t surpassed humans in mathematical abilities, which is a cornerstone of information science.
Most individuals forget that these AI models are literally a variety of model called large language models (LLMs), specifically designed to predict the following word from a pre-calculated probability distribution.
These models can only output, or base their output, on data they’ve seen; they will only go off what exists and never necessarily create anything “brand latest.”
The information science job requires developing novel solutions to unseen problems. In reality, we really need data scientists and machine learning practitioners to construct these AI models in the primary place and maintain them!
AI Still Makes Mistakes
As someone who works with these tools each day for a spread of applications, AI makes so many mistakes it’s ridiculous.
These LLMs often “hallucinate”, which is a term you’ve gotten likely heard and is when these AI models produce outputs that appear plausible but are literally very incorrect.
This stems from the proven fact that they’re probabilistic models by nature and may potentially “string” words together that make no sense to satisfy users’ demands or expectations.
Humans also make mistakes, however the difference is that the majority humans are aware of their mistakes after you correct them. They’re not uber-confident of their initial response either, depending on the scenario.
Whereas AI is sort of stubborn, clever, and really certain of the answers it gives you, which psychologically tricks us, humans, into pondering it’s correct.
Imagine how jarring this might be in a piece setting.
An AI data scientist wouldn’t give you the chance to accurately gauge how outrageous or ridiculous its output is, and so it fails to set expectations while you implement its’ given solution.
It misses that lack of nuance and intangibles us humans have about many data science and machine learning projects.
Limit To Performance
What’s interesting to me is that these AI models aren’t actually getting substantially higher over time.
The explanation is twofold:
For instance, OpenAI’s GPT models have been trained mainly on the entire of the web to a certain extent, there isn’t much “latest” data for it to make use of.
This data also comes from humans, so it may possibly’t exceed human intelligence; that’s its ceiling.
These AI models won’t get any higher unless there’s a large scientific breakthrough within the underlying algorithm.
And the proven fact that they won’t get any higher means the present state will remain the identical, and AI has not yet replaced data scientists.
Can’t Construct Relationships
AI is incapable of relationships, despite what number of individuals are sadly getting emotionally attached to those robots.
Humans are social creatures, and a lot of the world’s business interactions are done through relationships.
People do business, hire, and work with people they like, even in the event that they might not be probably the most “technically” qualified.
A stakeholder will trust you as an information scientist if you’ve gotten delivered consistent results for them.
Even when an AI comes up with a “higher” solution to their problem, the stakeholder will likely prioritise you as a consequence of the intangible human relationship you’ve gotten built.
Every job relies on human connection. Some parts shall be automated, but many won’t.
Within the case of an information scientist, it will be incredibly hard to automate:
Any lively human part could be inconceivable to interchange.
Has Anything Really Modified?
Considered one of my old line managers once asked me:
Sure, we now have higher tools to resolve certain problems, and productivity in certain elements of our jobs has increased, but the information scientist role truthfully hasn’t modified that much.
Take a minute and take into consideration what has materially modified in your day-to-day life from AI.
AI, in its current form, has been around for greater than 4 years, yet society as an entire hasn’t been significantly impacted from where I’m standing.
That’s all that should be said here.
If, after reading this, you actually need to dive deep into learning AI, I like to recommend my previous post, which provides you a full, in-depth roadmap of every part it’s essential to master AI.
You possibly can test it out below!
One other Thing!
Join my free newsletter where I share weekly suggestions, insights, and advice from my experience as a practising data scientist and machine learning engineer. Plus, as a subscriber, you’ll get my FREE Resume Template!
Dishing The Data
newsletter.egorhowell.com
