🦥 The Future of Software Engineering Interviews

Sponsored by

Read on our website

Hello humans,

Welcome to this week’s Sloth Bytes. I hope you had a great week.

Poll results!

Looks like a majority would like/don’t mind a second email!

I’ll test it out next week. I’ll be sending an email on Tuesday and Thursday.

You can give me your thoughts Thursday👀 

Free, private email that puts your privacy first

A private inbox doesn’t have to come with a price tag—or a catch. Proton Mail’s free plan gives you the privacy and security you expect, without selling your data or showing you ads.

Built by scientists and privacy advocates, Proton Mail uses end-to-end encryption to keep your conversations secure. No scanning. No targeting. No creepy promotions.

With Proton, you’re not the product — you’re in control.

Start for free. Upgrade anytime. Stay private always.

Get free private email

Go from AI overwhelmed to AI savvy professional

AI will eliminate 300 million jobs in the next 5 years.

Yours doesn't have to be one of them.

Here's how to future-proof your career:

• Join the Superhuman AI newsletter - read by 1M+ professionals

• Learn AI skills in 3 mins a day

• Become the AI expert on your team

Start learning AI now

The Future of Software Engineering Interviews

A lot of you were interested in a job market/interview process, so I did some research!

And uh… it's weird.

The Great Interview Split

Right now there's basically two sides of tech interviews:

Side 1: The traditional process. LeetCode problems, system design, behavioral questions, no AI tools allowed. Maybe you can use Google, but that's it.

Side 2: Forget the traditional process, you can use AI tools during your interview.

Let's talk about both.

Side 1: Big Tech Isn't Changing (But Getting Harder)

Big companies including FAANG (or MANGA, or whatever acronym we're using this week) are NOT changing how they interview.

They're still doing the same process:

• Technical LeetCode problems

• System design rounds

• Behavioral interviews

• Maybe some take-homes

But that doesn't mean it's all the same.

The bar is getting higher.

The bar has shifted approximately one standard deviation higher.

Which means an interview performance that got you hired in 2021 might not even pass screening today.

Here's what's changed:

1. The questions are getting harder

One engineer interviewed at Google in 2021 and again in 2024 noticed a big difference:

Yikes.

Be prepared for more LeetCode medium and hard problems as standard.

2. Your solution needs to be more robust

Interviewers now routinely expect:

• Proper error handling

• Robust input validation

• Clean, production-ready code

• All within the same time constraints as before (15-30 minutes per problem)

3. Companies are actively changing questions to combat AI cheating

AI is becoming a serious problem in interviews.

Because of this, Big Tech companies are designing questions that:

• Can't be easily copy-pasted into ChatGPT

• Require explaining your thinking process, not just the answer

• Have multi-step reasoning that trips up LLMs

• Include follow-ups that probe whether you understand the fundamentals

If you're studying, this means:

• Memorizing solutions won't cut it anymore. You need to understand WHY a solution works.

• Be ready to explain your thought process. Interviewers are listening for AI patterns vs. human reasoning.

• Practice variations, not just standard problems. The exact LeetCode question won't appear anymore.

4. System Design is more likely & has higher expectations

System design rounds are no longer just for senior engineers.

They're becoming standard for junior roles and even intern roles.

Not only that, but the expectations are higher.

One engineer with 15 years at Google re-entered the market and was shocked:

Companies now expect:

• Knowledge of specific modern technologies (Kafka, Redis, Kubernetes)

• Experience with distributed systems at scale

• Understanding of real-time data processing

• Familiarity with cloud architecture patterns

Even if you've never used these in your day job.

Other Changes

• Downleveling: offered a job at a lower level than the one applied for

• AI in system design: design LLM-based systems, chatbots, AI video apps.

• Behavioral questions ask about AI: "How do you use AI tools?" "Describe AI features you built."

• Team matching became a second interview: One staff engineer waited 4 months after passing technical rounds. By then, all competing offers expired.

• In-person interviews coming back

Side 2: The AI Revolution

Now this is where it gets interesting.

Some companies looked at their interview process and said

"Wait, our engineers use AI tools every day. Why are we testing people without them?"

Meta’s Approach

Meta is piloting coding interviews with full AI access. Candidates can use an assistant throughout the session, just like real work.

They don’t care if you memorize algorithms; they care if you can think, debug, and collaborate with AI.

Canva’s Approach

Canva straight-up expects candidates to use tools like Copilot, Cursor, or Claude during interviews.

Half their engineers already do daily. So instead of banning AI, they built interviews around it. They’re judging how you guide and improve AI output, not how well you avoid it.

Anthropic’s Approach

Anthropic encourages candidates to use Claude to prep and brainstorm, but not to fake it.

Their rule: use AI to refine your ideas, not replace them.

They value authenticity and judgment over automation.

Rippling’s Approach

Rippling allows candidates to use AI during coding rounds, but it’s optional.

They’ll evaluate you differently if you do use AI by focusing on how strategically you collaborate with AI, not whether you can brute-force every solution.

The “Vibe Coding” Era

Some startups have gone full “vibe coding.”

These interviews aren’t testing how fast you can type; they’re testing how you think with AI.

The goal isn’t perfection, it’s seeing how you recover when AI messes up.

Think interviews where you:

• Share your ChatGPT prompts live

• Build a todo app using Claude

• Design an API while screen-sharing in Cursor

What you should do to prepare?

• Prepare for all interview types

• Learn how to use AI right

• Focus on fundamentals (they still matter)

• Be prepared for AI or no AI interviews

Thanks to everyone who submitted!

NeoScripter, quentin0133, gcavelier, mkgp-dev, grcc492, and saleh-ato!

Nom Nom Numbers

A number can "eat" the number to its right if it’s larger than that number.

When it eats, it becomes the sum of both numbers.

Keep repeating this process from left to right until no more eating can happen.

Examples

nom_nom([5, 3, 7])
output = [15]

nom_nom([5, 3, 9])
output = [8, 9]

nom_nom([1, 2, 3])
output = [1, 2, 3]

nom_nom([2, 1, 3])
output = [3, 3]

nom_nom([8, 5, 9])
output = [22]

nom_nom([6, 5, 6, 100])
output = [17, 100]

How To Submit Answers

Reply with

• A link to your solution (github, twitter, personal blog, portfolio, replit, etc)

• or if you’re on the web version leave a comment!

• If you want to be mentioned here, I’d prefer if you sent a GitHub link or Replit!

That’s all from me!

Have a great week, be safe, make good choices, and have fun coding.

If I made a mistake or you have any questions, feel free to comment below or reply to the email!

See you all next week.

Want to advertise in Sloth Bytes?

If your company is interested in reaching an audience of developers and programming enthusiasts, you may want to advertise with us here.