Resume : AI applied for the job

The Most Over qualified Yet Under qualified Employee You’ll Ever Meet

Gunjan Titiya
5 min readOct 28, 2024
AI intern applying for the job

Hello AI leaders,

Have you ever had that one intern? The one who graduated top of their class, has three degrees and a PhD, yet you won’t be trusting them to give production access on day one? Meet AI, the digital version of that intern — except this one might make the same mistake at lightning speed, in 37 languages.

While we are closing in on 2024 in 60 days, there is no denying of the most important reality check of 2024: Artificial Intelligence. It changed the business landscape forever, delivered 10x productivity boost to those embraced it early on and also managed to attract wide skepticism (and rightfully so) of the cost of ‘Productivity gains’ is directly correlated to ‘Jobs lost’.

Despite all its incredible capabilities, is actually your least intelligent employee. And that’s exactly how it should be.

Let’s review Resume of AI so far, and we will assign it the right role that fits the best for our needs.

Resume of AI

ARTIFICIAL INTELLIGENCE

I can hold entire internet's knowledge in my brain

PROFESSIONAL SUMMARY

- Highly capable digital entity seeking to revolutionize your
business operations.
- Can process billions of calculations while simultaneously
writing poetry and predicting market trends.
- Never needs coffee breaks or complains about Mondays.

CORE COMPETENCIES
- Processing Speed: 1 million tasks per second
- Languages Known: All of them (including some that don't exist)
- Memory: Literally everything on the internet
(up until the date I was trained)
- Availability: 24/7/365

ACHIEVEMENTS

- Beat world champions at chess , bar exam, SATs and all other
competency exams- all at once.
- Generated 10,000 marketing copy in the time it took you to read this
- Created art that definitely looks like... something
- Talk (almost) like a human with any personality you want me to take on

WHAT I CAN DO FOR YOU

- Automate your entire sales, marketing, CRM systems and
get leads for you while you sleep
- I talk to your customer impersonating you (umm...)
in form of chatbot, phone caller, video caller,
or be a podcast guest for you.
- Generate product photography for your eCommerce store without
even hiring a photographer.
- Basically anything you measly humans don't like to do

Impressive, I must say, but let’s read the performance review of the AI from the business leaders who “hired” AI this year.

The Reality ( Performance review of AI)

THINGS IT ACTUALLY DOES WELL:

The Data Powerhouse

  • Processes vast amounts of structured data accurately
  • Spots patterns humans might miss in large datasets
  • Generates basic reports at impressive speed

Great as an Assistant

  • Drafts decent first versions of routine documents
  • Summarizes long content effectively
  • Handles basic email classifications and routing

Some analytical tasks

  • Quickly identifies trends in sales data
  • Flags unusual patterns in transactions
  • Runs multiple scenario analyses simultaneously

THE REAL LIMITATIONS:

The Planning Paradox

  • Confidently suggests completing 6-month projects in a week
  • Creates project timelines assuming everyone works 24/7
  • Completely ignores human factors like meetings, breaks, and sleep

The Context ? Error 404.

  • Suggests marketing luxury cars to teenagers with no income
  • Creates support responses that answer the question perfectly but miss the actual problem
  • Recommends business expansion to markets that don’t exist

The Judgment Gap

  • Equal confidence in both brilliant and disastrous suggestions
  • Proposes budget cuts in critical areas based purely on numbers
  • Generates leads without understanding qualification criteria

AREAS FOR IMPROVEMENT (With Human Supervision):

Strategy Development

Human Needed For:

  • Understanding market context
  • Evaluating realistic timelines
  • Assessing resource constraints
  • Identifying potential risks

Content Creation

Human Needed For:

  • Brand voice consistency
  • Cultural sensitivity checks
  • Fact verification
  • Context-appropriate messaging

Customer Interaction

Human Needed For:

  • Handling complex customer issues
  • Understanding emotional subtext
  • Managing escalations
  • Building real relationships

ACTUAL BEST USE CASES:

1. Augmentation (Not Replacement)

  • Supporting human decision-making with data
  • Providing first drafts for human refinement
  • Handling routine tasks to free up human creativity
  • Processing and organizing information for human analysis

2. Efficiency Enhancement

  • Automating repetitive processes
  • Streamlining document workflows
  • Providing quick data insights
  • Supporting (not replacing) customer service

3. Quality Improvement

  • Catching human errors in data entry
  • Ensuring consistency in documentation
  • Maintaining regular communication schedules
  • Standardizing routine processes

PERFORMANCE IMPROVEMENT PLAN:

Clear Boundaries

  • Stick to data processing and analysis
  • Leave strategic decisions to humans
  • Focus on augmentation, not replacement

Supervision Requirements

  • Human review of all customer-facing content
  • Regular audits of AI-generated suggestions
  • Oversight of resource allocation recommendations

Success Metrics

  • Reduction in routine task time
  • Accuracy of data processing
  • Quality of first drafts
  • Time saved for human workers

The Beautiful Symphony of Dumb

Remember that scene in “The Office” where Michael Scott drives his car into a lake because his GPS told him to? That’s AI in a nutshell. Incredibly powerful, remarkably capable, and absolutely clueless about whether driving into a lake is a good idea.

But here’s the thing: We don’t need AI to be smart. We need it to be reliable, consistent, and really, really good at boring stuff.

The Perfect Partnership

Think of it this way:

  • You’re the chef who knows how flavors work together
  • AI is your tireless sous chef who can chop onions perfectly for 12 hours straight without crying

Or in business terms:

  • You’re the strategist who understands market nuances
  • AI is your analyst who can process market data faster than you can say “blockchain”

Why This Is Actually Great News

Remember the panic when calculators were introduced? Teachers worried no one would learn math anymore. Instead, we got better at complex problem-solving because we weren’t busy doing manual calculations.

The same thing is happening with AI. It’s not making humans obsolete; it’s making us more human.

The “Would I Trust This With My Intern?” Test

Want a foolproof way to know when to use AI? Ask yourself: “Would I trust this task with my most brilliant yet literal-minded intern?”

✅ Organizing 10,000 customer feedback responses into categories

❌ Deciding how to respond to an angry customer

✅ Generating 50 social media post ideas

❌ Deciding which ones won’t accidentally offend someone

✅ Analyzing market trends

❌ Understanding why those trends matter

Building Ethical AI: The Path Forward

Here’s the real talk: AI isn’t the smartest thing ever created — it’s the most powerful calculator we’ve ever built. And that’s okay. In fact, it’s preferable.

Why “Dumb” AI is Better:

  1. It’s predictable
  2. It’s controllable
  3. It knows its limitations (or at least, we do)
  4. It forces us to maintain human oversight
  5. It keeps us honest about what technology can and cannot do

The Liberation of Lower Expectations

Stop trying to make AI think. Let it thunk instead.

  • Thinking: Complex reasoning, understanding context, making nuanced decisions
  • Thunking: Processing, calculating, organizing, generating based on patterns

When we embrace AI for what it is — a powerful tool rather than an artificial mind — we can finally use it effectively.

So let’s stop worrying about AI becoming too smart and start focusing on making it more useful for your business. After all, the best employee isn’t always the smartest — it’s the one who makes your job easier while keeping you entertained with their occasional brilliant mistakes.

until next time.

Gunjan

This article is proudly brought to you by Bytes & Bayes, we build Production grade AI products from scratch. Book a call with us to evalute your data needs and how we can help!

--

--

Gunjan Titiya
Gunjan Titiya

Written by Gunjan Titiya

Founder Bytes & Bayes | Data and AI strategy consultant | Speaking, Writing about data world | "LLM models wont improve company's bottom line, their Data will."

No responses yet