Last week, I hosted the first live Q&A of 2026 across LinkedIn, YouTube, and Instagram. It was exciting to meet everyone joining from all over the world!These live sessions exist for one simple reason. I get far more thoughtful questions than I can reasonably answer one on one. The format allows me to address patterns I see repeatedly.This newsletter summarizes the most important themes and answers from that session.
Exaltitude newsletter is packed with advice for navigating your engineering career journey successfully. Sign up to stay tuned!
One of the most common questions was this:
What kind of experiences can junior engineers take to bridge the gap between graduation and more established roles?
This is a difficult market for junior engineers. There is no way to sugarcoat that. The strongest signal you can have when you graduate is not coursework, certificates, or personal projects. It is internships.
Here is the uncomfortable truth.
Your junior year internship largely determines your full time entry level job options.
That means the real work starts earlier than most students expect.
If you wait until junior year to think about internships, you are already competing at the hardest stage. The students who do best tend to follow this progression:
Freshman and sophomore internships are harder to find and often unpaid. They still matter. They are what make you competitive later.
If you graduate without internships, employers have no signal. They cannot see how you operate in a professional environment. That is why the market feels brutal to many new grads. The system is not designed to evaluate potential. It evaluates signals.
For more tips on internships, watch Software Engineering Internships: 7 Things They Don't Tell You.
Another question stood out because it reflects a widespread anxiety.
I want to learn AI, but I cannot learn it the same way as before. I need to catch up fast. What do you suggest?
This question is based on a false assumption.
It is not too late to learn AI.
Most people working in AI today did not start with AI. They started with programming, math, or data. Many are still learning fundamentals.
Trying to skip foundations does not save time. It wastes it.
If you want to work in AI or machine learning, you must understand:
Jumping straight to tools without foundations leads to a shallow understanding. You may be able to demo something, but you will not be able to debug, adapt, or go deeper. That is where people stall.
There is no shortcut that replaces fundamentals. The good news is that you are not behind. You can learn this step by step, even now.
I have free roadmaps for AI engineering, data science, Python, and math. They exist precisely because the noise is overwhelming. Structure matters more than speed.
Yes. Unequivocally.
Software engineering is not disappearing. It is becoming more foundational.
Even when you are not building models, strong engineering fundamentals determine whether AI products work in real environments. Software engineering skills transfer across AI, backend systems, infrastructure, and product development.
In 2026, engineering fundamentals are still one of the safest skill investments you can make.
Another recurring concern was this:
Frameworks and tools change so fast. How do you stay current without feeling overwhelmed?
The mistake is trying to learn everything.
Engineers in industry do not learn randomly. They learn toward a goal.
When you are working, your priorities are defined by what you are building. As a learner, you do not have that structure, so everything feels equally important. It is not.
You need to define a target role.
Examples:
Once you define the role, learning becomes selective. You learn what serves that role and ignore what does not.
Some skills overlap across roles. Python is one of them. Others are highly specialized. Cybersecurity certifications, for example, matter only in specific contexts.
You cannot chase every new framework. No one can. Pick a direction, then go deep enough to be credible.
To explore different roles, skills required for each, and median pay, watch "What Programming Languages To Learn In 2026"
One question cut closer to engineering judgment:
How do you tell when AI is improving engineering quality versus just helping teams ship faster?
Quality should never be optional.
The best outcome is shipping faster while maintaining high quality. Any organization that prioritizes speed at the expense of quality is building long term risk into its systems.
Strong engineering cultures do not reward output alone. They reward correctness, maintainability, and long term thinking.
AI should amplify good engineering judgment, not replace it.
If a team ships quickly but accumulates fragile systems, the cost shows up later. Usually painfully.
For tips on best coding practices while using AI, watch "99% of Engineers Are Using AI Tools Wrong."
Several people asked how to choose between data focused roles and AI roles.
The distinction matters.
Neither is better. They suit different preferences.
If you are unsure, watch "AI, Machine Learning, Data Science: Which is the Better Career."
Short answer: usually no.
Certificates matter only when they are explicitly required in job descriptions. Cybersecurity is a common example. Some regulated roles also require them.
For AI, machine learning, and general software engineering in the US, certificates rarely influence hiring decisions.
Certificates are useful only if they help you learn something you could not structure on your own. They are not a hiring shortcut.
Always check job postings in your target market. If certificates are not listed, they are not the signal employers are looking for.
Historically, dominant tech companies emerged due to network effects. Messaging apps are the classic example. You use the app your friends use.
AI does not work that way.
You do not need your friends to use the same model for it to be useful. That removes the moat. Models can be copied, improved, or leapfrogged quickly.
That is why no single LLM has durable dominance today. Quality alone is not enough to lock in users.
This dynamic is fundamentally different from previous consumer tech waves.
The final question was practical and important.
If you had one month off and intermediate AI skills, how would you build your portfolio?
I would work on a real problem for someone else.
I'd offer my skills/time to a startup, a founder, or a team building something real. Even unpaid work can become professional experience if it solves an actual need.
Personal projects are hobbies. Projects built for real users are professional experience.
That distinction matters in interviews and job search.
If you are following along on this journey, thank you. We just passed 50,000 on LinkedIn and are closing in on 100,000 on YouTube. I appreciate every single person who shows up, asks real questions, and thinks seriously about their future.
See you at the next live!
I’ll be hosting another live session in February!
[LinkedIn] Reserve your spot here.
[YouTube] Sign up here.
If you want your question included in the next session, leave a comment on the events!
Exaltitude newsletter is packed with advice for navigating your engineering career journey successfully. Sign up to stay tuned!
.png)
