Emerging Tech Skills not Displaced by AI

 |   |  Disclaimer: Links to some products earn us a commission

4 months ago, IBM replaced 200+ Human resource workers with AI agents[1] that automated many of the workflows. However, this automation also created new tech jobs, and overall, the mass HR lay-off co-occurred with more hires.

Arvind Krishna, CEO of IBM, told The Wall Street Journal that “While we have done a huge amount of work inside IBM on leveraging AI and automation on certain enterprise workflows, our total employment has actually gone up, because what it does is it gives you more investment to put into other areas.”

These new jobs are in critical thinking areas that involve sales and engineering, especially work that already has AI involved. These jobs were awarded to people who demonstrated intellectual value.

4% of Microsoft’s workforce, a total of 9000 workers, were marked in July 2025[2] for a mass layoff. This is one of many rounds of layoffs, previously being another 6000 in May. All of this is to refocus their resources efficiently, with AI assisting most jobs. Now, 30% of the company’s code is reportedly written by AI, and employees are encouraged to use AI.

This tells a 2-tiered story – fewer people, more productivity. The unskilled and no-longer effective ones are likely to be laid off. The unskilled individuals are under threat because companies save money by eliminating them. However, those who no longer have a valid scope in a company are also getting laid off. Their skills are no longer needed or useful, even though they weren’t unskilled (like the HR workers at IBM).

Top-level summary

  1. AI displaces jobs in big companies, but also creates more unique jobs.
  2. 10,000 AI manager job postings in India suggest new skills for managing AI tools.
  3. Trending “vibe coder cleanup specialist” job title on LinkedIn suggests engineers who clean up AI code are becoming marketable, skilled professionals.
  4. Diversity of AI tools and technologies leads to a gap between those who create technology and those who use technology to create something new.
  5. Human experiences can still, in real-time, outwit GPT users.

AI janitors and AI managers

Another new development is the AI vibe coder cleanup specialist job title that’s going viral among devs on LinkedIn. As the title says, these are really smart engineers who understand the big picture and have the technical skills to fix the code written by weaker coders who predominantly copy-paste from LLMs without understanding the code (vibe coders). This fixing requires a lot more than coding skills.

  1. It needs an understanding of the purpose of coding a certain module.
  2. It needs a clear understanding of what state variables and computations are necessary (essentially, is this function necessary, and do we need to process this?).
  3. It needs a high comprehension of syntax and refactoring, and also attention to detail for work that does not seem intuitive.

There are many other such jobs – AI translation reviewer, Prompt evaluator, AI text cleanup specialist, etc. All of these are jobs that are currently open because AI has entered the workforce. However, even these jobs are not resistant to job displacement because one of the main goals of most AI companies is to create better AI. So, the number of humans with the intellectual capacity to beat AI will keep dropping while AI gets better. These professionals will be fewer in number, but probably fetch more money since it’s a high-value skill – to be better than AI at your task.

AI managers, those who implement AI strategies – for either hype or actual productivity – are a new breed of workers who generally have experience using the new tools and make educated decisions about how to implement AI. However, this job remains dispersed in the workforce and hasn’t taken shape because the successful ones are likely to already have technical insight that can’t be picked up through reading AI-related content.

Indeed India[3] lists about 10,000 AI manager jobs that generally require some technical expertise with the job description of managing AI platforms and tools used within a company, and also recommend new workflows using other AI tools.

The trend is that many older work tasks are being done by AI, which leads to a new scope for what humans can do. That is, AI is not just taking old jobs; it is creating new types of jobs.

Bridging skills

One type of skills category emerges directly from the high number of new technologies that keep coming up every day. These are the “bridging skills”.

I’m seeing a massive variability in people’s skills now. Many younger students are getting good at using all of the tech built by the older generation. You’ve probably seen that too.

If you are online, you’ve seen people who have launched new companies, new apps, new sites, new client services, etc., because AI-enabled them to do so. These are people who moved fast and built something. However, they didn’t build the technology itself – that happened in academia and R&D departments in big companies. Both these sides are made up of fairly exclusive groups of people. 2 sides – appliers & creators

The existence of these groups of people creates a new space for high-value skills:

Skills that bridge the fundamentals of innovation and applications using that innovation.


🧬 Creators – Innovators in the fundamentals of tech (Innovation)
Experienced people are in a position to build incredible technology through their skills and position in the industry. This is their unique value.

These people theorized neural nets, machine learning, created GPUs and NPUs, optimized data transfer protocols, etc. They created the LLMs and the documentation to use them.

🦴 Appliers – Early adopters of new tech (Application)
The younger ones are in a position to start at a new baseline created by the older generation. Their push toward new use cases, answering new problem statements, is unique and highly useful, but orthogonal to the innovation done by the big tech companies.

These people built businesses and apps using their creativity to use technologies created by the creators. They used those tools to automate a workflow at their job. They are power users of technology.

As both these groups push their own boundaries, the gap between creating and applying increases. More people specialize simply because the space in which either creation or application occurs expands exponentially.

This gap needs to be filled by newly skilled professionals who can bridge the innovation that is being created through OpenAI, Cursor, Microsoft, etc., and the applications being built through their R&D.

These newly skilled people will eventually become very valuable if they have these “bridging skills”

Help me run this site with a donation :)

The skills:

1. System architecture, old and new technologies
2. Hands-on experience trying out new innovations, new tools
3. Read research studies and patents
4. Connect between labs and industry
5. Have a home set-up in which they experiment
6. Engage the community of other innovators
7. Aware of new trends
8. Identify problems and attempt to solve them at their own expense
9. Creators of applications, open-source contributors
10. Mathematical thinking & execution
11. Framing human behavior

These people will tend to seamlessly connect those working at the fundamentals and those working at the application level.



Both groups (creators & appliers) are going to find it hard to converge, unless someone helps them converge.

New professionals will bridge the gap by communicating and moderating information flow between both groups. But communicating that requires knowledge of both their “vocabularies” and approaches. So, this new area of skills is also technical, while it is self-motivated and unstructured.

Human speed skills

While we discover new types of jobs as the industries evolve, I would highlight one more type of person who has that special, unique human value: Humans who outwit other humans that can’t get anything done without GPT.

It’s one of the oldest skills – experience + memory + speed

Some people will ideate and solve problems faster than most people can prompt an AI to do the same. This human quality will be one of the biggest unfair advantages.


Consider a person who spends 200 milliseconds to 5 seconds to say something factual and solve problems while having a conversation.

Vs.

Hours of stressful AI prompting after a meeting to get the same answers and construct a solution after the conversation, and calling more meetings to present their ideas, only to iterate over and over again.

First is the Knowledgeable Brain.
Second is the knowledge-deprived Brain + AI.

The first one will always prevail over the second when the time pressure is brutal and opportunities to shine are scarce.

These tech and non-tech workers possess the skills to have information readily available in their heads, along with improvisational and critical thinking skills, as well as numerous other cognitive skills and personal experiences that benefit them and demonstrate confidence in their abilities.

Human skill and human learning have never been more important, even if AI can give you knowledge. This speed of thinking with learned Knowledge in real-time is a premium humans have over AI-augmented humans.

Rate the article

Average rating 5 / 5. Vote count: 1

We are sorry that this post was not useful for you!

Let us improve this post!

Tell us how we can improve this post?

Sources

Check out these quick visual stories
Previous

Why split-screen ASMR Reels are so engaging

Join 3,749 other subscribers

Comments

Discover more from Cognition Today

Subscribe now to keep reading and get access to the full archive.

Continue reading

6 Forms of Beauty: Which do you admire? Speed changed our Attention span 6 Skills You are better at than AI How does Cognition happen? How to Biohack your Energy Level (7 steps) 6 Hobbies that teach Psychology Mental Effects of 7 mundane Daily Habits 7 Test-Anxiety Solutions, very easy Why we Fall for Misinformation so Easily 9 Cognitive biases affecting most of our decisions Your FYP affects mental health: Hints from 40 studies New mental health Ideas that challenge Old psychology Weird Quirks of the English Language These 7 study habits are harming your grade 8 Sentences to Rewire the Mind for Mental Health Big Braining: Top 9 Hobbies to refine Critical Thinking 6 Best Study techniques for Memorizing Facts Mind-Body Connection: 8 hobby ideas for 2025 6 Universal Words across most Languages Neural Circuits of Productivity Uncovered Gain fluency in a new language at record speeds (How-to) Core Financial Concepts in Psychology Why Music is so Important, psychologically DIY Upskilling: 10 psychological insights Seeing a face? It’s Pareidolia 10 Lesser-known Psychology Facts you didn’t know 8 Types of Memory Systems the Brain has by default 8 methods of learning that make you a fast learner ChatGPT 4o shows 7 unique cognitive biases 9 Learning Science Principles for Educators