I keep hearing over and over that it is important in one's profession to have "AI skills". I understand how AI is going to change the work force in the future, but what particular AI skills are important to have?
So far I am a SME with AI computer hardware, networking and infrastructure. I have built an agent at home, used few shot tuning on a generic LLM, and did a resume builder and mock interview on it. I have a small prompt library that I keep at work and at home, slowly expanding to making skills and eventually larger tools. Is this enough to say "I have AI skills"? Is this going to become the same thing as saying "I have MS office and internet skills" in the future?
Should I be focused on training, prompting, building, RAG, embeddings, ethics, coding, APIs, genAI, model tuning, agentic workflows, and certifications before I say I have AI skills?
Basically which skills are desired or required in the future?
8 answers
Kelly Fry, MBA, PMP
Kelly’s Answer
The first step is awareness and understanding. It appears you know this well. On your resume, you should list the specific tools/models that you like to use and that you are familiar with prompting, etc.
Second (and this is where you can differentiate for now) is how are you using the tools to add value -- make money or save money. Even if you are able to streamline workflows, that results into time saved = dollars retained.
Once you are able to use your knowledge to create value from the tools you understand, you will be in the top 5% or 1% of AI users and this will become a key competitive advantage. When discussing this on your resume, be sure to lay out the quantifiable benefit (numbers) to showcase how these tools and skills have definitive value.
Joseph’s Answer
AI skills are similar. Increasingly, "I have AI skills" is meaning at the bottom end - "I can get ChatGPT to write an email for me" or "I use Gemini for search" - which is increasingly becoming an expectation rather than a skill above the rest. However, what you've described is definitely more of the advanced skill set, certainly around the LLM side of AI. (you didn't list much around traditional ML "AI", but you're allowed to be a specialist in one form of AI and still claim "Advanced AI skills".)
I'd probably recommend you provide clarity in CVs/applications on what you mean by "AI skills" - rather than just "AI skills", say something like "Advanced AI skills including model tuning, agentic workflows and prompt engineering" - from what you've described, I think that's something you can legitimately claim.
Audrey’s Answer
The most important thing is how willing you are to adopt AI and incorporating it into your life so that you can display this in an interview for a job. If used correctly you can minimise the time you spend on 'busy-work' and optimise the time spent on high level / priority tasks.
Hope that helps!
Hemant’s Answer
- Show value, not just knowledge — the real differentiator is "I used AI to save X hours or automate Y task" not just "I know how to prompt"
- Put specific tools on your resume — LangChain, specific LLMs, vector DBs, whatever you've touched — name them explicitly
- Add RAG next — it's the #1 skill companies need right now for working with their own internal data
Learn MCP (Model Context Protocol) — Anthropic's open standard for connecting AI agents to real tools, files, APIs — this is becoming the backbone of agentic systems, learn it now before everyone else does
- Look into ACP (Agent Communication Protocol) — emerging standard for how agents talk to each other in multi-agent systems, very early but very important
- Sandboxes & safe execution — knowing how to run AI-generated code safely (E2B, Docker sandboxes, isolated environments) is a real skill as agentic systems grow
- API + workflow integration — connecting AI to real systems (Slack, databases, internal tools) is where the money is
- Even basic Python — you don't need to be a dev, just enough to read and tweak scripts
- Quantify everything — "reduced X process from 2 hours to 10 minutes using an agent I built" beats any certification
On the MS Office comparison — you're right, it's coming:
- Basic AI use will be table stakes soon
- Building and integrating AI systems will stay rare and valuable for a long time
Your hardware + infrastructure background is a secret weapon — most AI people can't do what you do on the infrastructure side. That combo is genuinely hard to find.
Keep building real projects. Real things you've built will always speak louder than any certification. You're on the right track — just keep going.
Adam’s Answer
However, there is much greater value in saying I have implemented AI to perform this task and this is how I solved this problem with AI tools. You are on the right track with certifications and additional professional development. Never stop learning new things because this will ultimately set you up for success when you become the subject matter expert for a new tool just because you were curious and played around with it.
Continuous improvement in all things you do will open doors and you'll be prepared to seize opportunity when its presented.
Sandeep’s Answer
You already have more AI experience than many people who simply list “AI skills” on a resume. The important part is not just using AI tools, but understanding how to apply them to solve real problems.
In the future, valuable AI skills will likely include:
* Working with APIs and automation
* Prompting and workflow design
* Understanding data and model limitations
* Building AI-assisted applications
Supreeti’s Answer
1. Layer 1 : Foundations (you already have most of this) - Prompting (beyond basics), understanding LLM behavior (hallucinations, context limits), basic agent workflows
2. Layer 2: Systems thinking (RAG, database etc)
3. Layer 3: Integration -APIs (must-have), Connecting AI to real tools (Slack, CRM, databases), Automation pipelines, Backend logic
Push yourself and build on what you have already learned and focus on orchestrating workflows based on efficiency and tool usage. Pick a layer and integrate systems that will help you or the organization reduce manual effort while utilizing minimum tokens. In the process you are designing, validating implementing, validating and debugging the system from end to end. Showing impact that will stand you out. If you are interested in security or specialization in a field then add certifications.
Natig’s Answer
The most sought-after skills in the future will not just be the ability to ask questions, but the ability to build systems. You should pay special attention to RAG (Retrieval-Augmented Generation) technology; because the most important thing for companies is to get AI to work properly with their internal, confidential data. In addition, learning Agent workflows (the coordinated work of several bots) and API integrations will greatly strengthen you from a technical perspective.
As for coding, even if you are not a professional programmer, understanding scripts written in Python and modifying them according to your needs will put you one step ahead of others. Ethics and security are the "responsibility" side of things, which are absolutely required in large projects.
In short, you've already laid the foundation. Now, if you focus on connecting these skills (deployment and workflow), you'll have solidified your place as a future expert. Certifications are great, but real projects you've built and agents working effectively will always speak louder.