Can’t Hurt Me -- The Ten Challenges That Can Transform Your Life

See other Biographies & Autobiographies

This blog captures the core philosophy and ten life challenges from David Goggins’s book Can't Hurt Me—not as motivation, but as a system for self-mastery.

This is not about hype.
This is about brutal honesty, discipline, and mental toughness.

---

1️⃣ Accountability Challenge – Face Your Past Honestly

What it means
Everyone has suffered—abuse, failure, poverty, rejection, fear. Most people bury it. Goggins says: write it all down.

Why it matters
Denial is comfort. Truth is power. Until you face what happened to you, you remain controlled by it.

Action step

• Take a notebook (not digital)

• Write every painful memory, excuse, and limitation

• Do not filter or justify—just tell the truth

This is where real change begins.

---

2️⃣ The Accountability Mirror – Stop Lying to Yourself

What it means
Put sticky notes on a mirror listing your flaws, weaknesses, goals, and insecurities.

Why it matters
The mirror does not care about excuses. It forces self-responsibility.

Action step

• Write statements like:

  • “I am overweight.”

  • “I am undisciplined.”

• Accept them without self-pity

• Add: “This can be fixed.”

Discipline is born here—not motivation.

---

3️⃣ The Discomfort Challenge – Leave the Comfort Zone Daily

What it means
Do at least one thing every day that feels uncomfortable but is good for you.

Why it matters
Comfort weakens the mind. Growth only happens in discomfort.

Examples

• Wake up earlier

• Exercise when tired

• Study after work

• Do the task you keep avoiding

Start small—but do it daily.

---

4️⃣ Callous the Mind – Make Pain Your Teacher

What it means
Just like hands develop calluses, your mind must develop resistance to pain.

Why it matters
Life will hurt you anyway. Training through pain prepares you to handle it.

Action step

• When your mind says “I’m done”

• Push a little further

• That extra effort builds mental armor

---

5️⃣ Taking Souls – Win Through Excellence

What it means
Instead of arguing, complaining, or proving yourself verbally—outperform everyone.

Why it matters
Excellence silences critics. It breaks opponents psychologically.

Action step

• Do your work so well that it cannot be ignored

• Let results speak

• Stay quiet, stay consistent

This is mental warfare—won through discipline.

---

6️⃣ The Cookie Jar – Store Your Past Wins

What it means
Create a mental “cookie jar” filled with moments when you overcame pain or failure.

Why it matters
When current suffering hits, past victories become fuel.

Action step

• Write down:

  • Times you didn’t quit

  • Struggles you survived

  • Habits you changed

• Revisit this list during hard moments

---

7️⃣ The 40% Rule – You’re Not Actually Done

What it means
When your mind says you are finished, you are usually only 40% exhausted.

Why it matters
The brain tries to protect you—not push you.

Action step

• When pain hits:

  • Run a little longer

  • Study a little more

  • Push one more rep

That remaining 60% is where growth lives.

---

8️⃣ Schedule Your Life – Time Is Not the Problem

What it means
Most people waste time and blame lack of opportunity. Goggins demands a full life audit.

Why it matters
Structure creates freedom. Chaos creates excuses.

Action step

• Plan your day in 15–30 minute blocks

• Track:

  • Work

  • Exercise

  • Learning

  • Rest

• Cut social media and mindless habits

Time must be earned, not assumed.

---

9️⃣ Never Settle – Greatness Must Be Maintained

What it means
Reaching a level is not success. Staying hungry is.

Why it matters
Complacency destroys excellence faster than failure.

Action step

• After every win, set a harder goal

• Never allow “good enough”

• Reset to zero and rebuild

Greatness is a daily decision.

---

🔟 After Action Reports – Learn From Failure Ruthlessly

What it means
Every failure deserves analysis, not emotion.

Why it matters
Most people fail repeatedly because they never study their mistakes.

Action step
Write after every setback:

1. What went right

2. What went wrong

3. What I could control

4. What I will change next time

5. When I will try again

This turns failure into a weapon.

---

🔚 Final Thought: You Are the Hero of This Story

This is not just David Goggins’ story.
It is a framework for ordinary people to build uncommon lives.

You don’t need to run ultramarathons or join the military.

You only need to:

• Stop lying to yourself

• Embrace discomfort

• Take responsibility

• Refuse to quit

The real question is:
👉 Are you ready to go to war with your own limitations?

Minimum operations (Medium Complexity) - Problem From HackerEarth

Index of "Algorithms: Design and Analysis"

Can you explain the solution to this problem:

Problem
You are given an array A of size N. You can perform the following operation on array A:

Select two indices i and j such that 1 ≤  i, j ≤  N. (note that i and j can be equal)
Assign Ai = Ai + 2
Then assign Aj = Aj - 1
You need to make all the elements of the array equal to zero.

Task
Determine the minimum number of operations required to make all the elements of A equal to zero. Else, print -1 if it is not possible to do so.

Function description

Complete the function solve() provided in the editor. This function takes the following parameters and returns the required answer:

N: Represents the size of the array A
A: Represents the elements of array A.

Input format

Note: This is the input format that you must use to provide custom input (available above the Compile and Test button).

The first line contains T, denoting the number of test cases. T also specifies the number of times you have to run the solve() function on a different set of inputs.
For each test case:
The first line contains N, denoting the size of array A.
The second line contains space-separated values, denoting the elements of array A.
Output format For each test case, print the output on a new line. Either the minimum number of operations required to make all the elements of A equal to zero or print -1 if it is impossible to do so.


Sample Input
2
1
-2
2
1 -1
Sample Output
2
-1
Time Limit: 2
Memory Limit: 256
Source Limit:
Explanation
The first line contains the number of test cases, T = 2.

The first test case

Given

N = 1
A = [-2]
Approach

We can perform the following operations on A:

Select i = 1, j = 1, A1 = -2 + 2 = 0 and A2 = 0 - 1 = -1, A becomes [-1].
Select i = 1, j = 1, A1 = -1 + 2 = 1 and A2 = 1 - 1 = 0, A becomes [0].
Hence, all the elements are equal to zero after 2 operations. Thus, the answer is 2.

The second test case

Given

N = 2
A = [1, -1]
Approach

It can be shown that it is not possible to reduce all the elements of A equal to zero simultaneously. Thus, the answer is -1.

Download Test Cases
Solve on HackerEarth

This is a great problem because it looks simple, but it’s really about invariants and feasibility, not simulation.

Let’s break it down cleanly.

---

🔁 The operation (very important)

In one operation:

• Pick indices i and j (they can be the same)

• Do:

  • A[i] += 2

  • A[j] -= 1

So each operation changes the array in a fixed way.

---

🔑 Step 1: Understand what never changes (invariants)

🔹 Invariant 1: Total sum increases by +1

Each operation does:

diff
+2  and  -1  → net change = +1

So after k operations:

sql
sum(final) = sum(initial) + k

But your goal is:

sql
final array = [0, 0, ..., 0]
⇒ sum(final) = 0

Therefore:

bash
0 = sum(A) + k
⇒ k = -sum(A)

📌 Conclusion 1

• If sum(A) > 0, then k < 0 → impossible

• So sum(A) must be ≤ 0

---

🔹 Invariant 2: What happens to individual elements?

An element:

• Can increase by 2

• Or decrease by 1

• Or both (if i == j, net +1)

So to reduce a positive value, you must apply -1 operations to it.
To increase a negative value, you use +2.

This creates a cost structure.

---

🔢 Step 2: What must the number of operations be?

From invariant 1:

ini
k = -sum(A)

So the number of operations is fixed, not something we can choose.

If a solution exists, the answer must be:

ini
answer = -sum(A)

No more, no less.

---

🧠 Step 3: Feasibility condition (the real trick)

Each operation gives:

• one -1

• one +2

Across k operations:

• Total negative capacity = k

• Total positive capacity = 2k

Now think element-wise:

For positive elements A[i] > 0

They must be reduced to zero using -1 operations:

bash
Total required decrements = sum of positive A[i]

But we only have k decrements available.

So we need:

nginx
sum_positive ≤ k

Substitute k = -sum(A):

scss
sum_positive ≤ -sum(A)

Since:

bash
sum(A) = sum_positive + sum_negative

This becomes:

scss
sum_positive ≤ -(sum_positive + sum_negative)
⇒ 2 * sum_positive ≤ -sum_negative

📌 This is the key feasibility condition

---

✅ Final Conditions for a Solution

A solution exists if and only if:

1. sum(A) ≤ 0

2. 2 * sum_positive ≤ -sum_negative

If both are satisfied:

bash
minimum operations = -sum(A)

Otherwise:

ini
answer = -1

---

🧪 Apply to the sample cases

✅ Case 1

makefile
A = [-2]
sum = -2 → k = 2
sum_positive = 0
sum_negative = -2

Check:
2 * 0 ≤ 2 ✔
Answer = 2

❌ Case 2

makefile
A = [1, -1]
sum = 0 → k = 0
sum_positive = 1
sum_negative = -1

Check:
2 * 1 ≤ 1 ❌
Answer = -1

---

🧠 What this problem teaches you

This is very important conceptually:

🧩 1. Don’t simulate when operations are large

Look for invariants instead.

🧩 2. Operations define resource constraints

Here:

• Each operation provides limited + and − power

🧩 3. Sometimes the answer is forced

The number of operations wasn’t something to minimize — it was fixed by math.

---

🧾 Final Pseudocode

python
sumA = sum(A)
if sumA > 0:
    return -1

sum_pos = sum(x for x in A if x > 0)
sum_neg = sum(x for x in A if x < 0)  # negative number

if 2 * sum_pos > -sum_neg:
    return -1

return -sumA

Failing Test Case

1
10
858 268 -73 -83 -613 -88 -240 -463 -605 -90

Tags: Algorithms,

Interviews For AI Engineer Role (2025 H2)

Index For Interviews Preparation    « Previously    Next »

1: Ericsson at Accenture

Q: What do you know about MCP?

Q: What is the difference between an MCP server and an API?

Q: Can we setup asynchronous connection with an MCP server? If yes, how?

### ### ### ### ### ### ### ### ### ### ### ### ### ### ### 

2: Macquarie (Australian Financial Firm)

Q: Talk about your overall experience. 

Q: Talk about your current/last project.

Q: What's prompting you to leave Accenture? What is that trigger?

Q: What do you know about Macquarie?

### ### ### ### ### ### ### ### ### ### ### ### ### ### ### 

3: abc@relx.com at a Legal Firm

Q: Can you explain the architecture of your current / last Agentic AI project?

Q: How would you evaluate a custom model that you have trained for classification purpose?

Q: Why and how did you choose framework for your Agentic application between LangGraph, LangChain, CrewAI, AutoGen, AI Refinery?

Q: Can you build a simple Agentic workflow with one agent and one or two tools (getWeather, getTraffic)?

4: Barclay's @ Accenture

Q1: I was asked about my AI/ML journey.
- Answer: Start with your latest project:
# Content Optimization platform in Ed-tech domain
# Then on top of that content, building mobile apps later donated to NGO in Educational cuase as part of CSR initiative of Accenture.

# Built three apps using vibe coding using Agentic Coding platforms such as Windsurf and GitHub Copilot.

In simple terms, I was generating content for the Ed-tech mobile apps first, seeding them in a SQLite database (so that the user has a local and offline access to the data), then vibe-coding the mobile apps.

... ... ... ... ...

Q2: Challenges faced in this project were:
- content relevance 
- grammatical correctness 
- complexity of content for different age groups

Q3: Explain your last project -- AI Over BI
- what it did
- how it worked
- talk a bit on the architecture

Q4: So you used LLM-As-A-Judge to reflect on the query generated by an LLM and produce some feedback. The question is: how do you know LLM-As-a-Judge is not hallucinating itself and that it is providing the desired feedback?

Q5: How does your application come out of generate-reflect-regenerate loop?
Answer: Two techniques to come out of loop:
1. Max iterations
2. Consistency check between last and second-last generated query.

Q6: How do you know that the two queries are consistent? What metric are you employing there?

Q7: Do you have a hands-on experience with 'procuring and setting up cloud resources' to 'building solution' to 'deployment on the cloud'?

Q8: If you observe that there is a latency issue with your RAG solution, what can you do about it?

Answer:
1. Trying out a faster LLM
2. Caching
3. Collecting component wise traces for checking which component is taking maximum time (viz usually the LLM but the traces will tell which is the second highest)

Q9: Which chunking strategy did you use and why?

Q10: Why didn't you use Semantic Chunking?

Q11: How did you prevent or reduce hallucinations in your RAG agent?

----- ----- ----- ----- ----- ----- ----- ----- ----- ----- ----- ----- 

5: IBM (Jan 2026)

Q1: Question on Pandas library around the functions such as merge(), groupby() and corr().

Q2: Have you built AI agents from scratch? If yes, using which framework?

Q3: Distinguish between AI Agent and AI Assistant.

Q4: Why does one-shot works better than zero-shot prompts?

Q5: Give an example where zero shot prompting seems to work just fine.

Q6: What is the structure of your current team?

Q7: Are you in a client facing role?

Q8: Describe a situation where you had to handle a client escalation?

Q9: Why are you leaving Accenture?

Q10: Why do you want to join IBM?

Index For Interviews Preparation    <<< Previously    Next >>> Tags: Interview Preparation,

Introducing Agentic AI (Chapter 1)

Download Book

<<< Previous Book Next Chapter >>>

From The Book: Agentic AI For Dummies (by Pam Baker)

What Is Agentic AI, Why It Matters, and How It Changes Everything

---

Introduction: Why This Chapter Exists at All

Let’s start with a simple observation.

If you’ve used tools like ChatGPT, Gemini, or Copilot, you already know that AI can:

• write text,

• answer questions,

• summarize information,

• generate code,

• sound intelligent.

But you’ve probably also noticed something else.

These systems don’t actually do anything on their own.

They wait.
They respond.
They stop.

Chapter 1 introduces a shift that changes this completely.

That shift is Agentic AI.

The core idea of this chapter is not technical—it’s conceptual:

AI is moving from “talking” to “acting.”

This chapter lays the foundation for the entire book by explaining:

• what Agentic AI really is,

• how it differs from regular AI and Generative AI,

• why reasoning and autonomy matter,

• why prompting is still critical,

• and how Agentic AI will reshape the internet and commerce.

---

The Simplest Definition of Agentic AI

Let’s strip away all the buzzwords.

Agentic AI is AI that can take initiative.

Instead of waiting for a human to tell it every single step, an Agentic AI system can:

• decide what to do next,

• plan multiple steps ahead,

• change its plan if something goes wrong,

• remember what it has already done,

• reflect on outcomes and improve.

In short:

Generative AI responds.
Agentic AI acts.

That one sentence captures the heart of the chapter.

---

Why “Agentic” Is Such an Important Word

The word agentic comes from agent.

An agent is something that:

• acts on your behalf,

• represents your interests,

• gets things done.

Humans hire agents all the time:

• travel agents,

• real estate agents,

• legal agents.

Agentic AI is meant to play a similar role—but digitally, and at scale.

It’s not just answering questions.
It’s figuring out how to solve problems for you.

---

Why This Is a Big Leap, Not a Small Upgrade

The chapter is very clear on one thing:

Agentic AI is not just “better chatbots.”

This is a qualitative change, not a quantitative one.

Earlier forms of AI:

• classify,

• predict,

• recommend.

Generative AI:

• creates text, images, and code,

• but still waits for instructions.

Agentic AI:

• decides what actions to take,

• sequences those actions,

• monitors progress,

• and adapts over time.

That’s a completely different category of system.

---

Agentic AI and the Road Toward AGI

The chapter places Agentic AI in a broader historical and future context.

What Is AGI?

AGI (Artificial General Intelligence) refers to AI that can:

• reason across many domains,

• learn new tasks without retraining,

• adapt like a human can.

We are not there yet.

But Agentic AI is described as:

a critical stepping stone toward AGI.

Why?

Because autonomy, planning, and reasoning are essential ingredients of general intelligence.

---

The Singularity (Briefly, and Carefully)

The chapter also mentions the idea of the technological singularity—a hypothetical future where AI improves itself so rapidly that society changes in unpredictable ways.

Importantly, the tone is cautious, not sensational.

Agentic AI:

• does not equal AGI,

• does not equal consciousness,

• does not equal sci-fi superintelligence.

But it moves us closer, which means:

• risks increase,

• responsibility increases,

• guardrails matter more.

---

Safeguards Are Not Optional

A very important part of this chapter is what it says must accompany Agentic AI.

Three safeguards are emphasized:

1. Alignment with human values
   AI systems must be trained and guided using objectives that reflect ethical and social norms.

2. Operational guardrails
   Clear boundaries defining what the AI can and cannot do—even when acting autonomously.

3. Human oversight
   Humans remain accountable for design, deployment, and monitoring.

This chapter makes one thing clear:

Autonomy without responsibility is dangerous.

---

Agentic AI Already Exists (Just Not Everywhere Yet)

One subtle but important point the chapter makes:

Agentic AI is not science fiction.
It already exists—just in limited, early forms.

Examples include:

• autonomous research assistants,

• supply chain optimization systems,

• multi-agent task managers,

• experimental tools like AutoGPT and BabyAGI.

These systems:

• plan,

• remember,

• coordinate tools,

• and operate over longer time horizons.

They are not human-like, but they are functionally useful.

---

Why People Are Afraid of Reasoning Machines

This chapter takes an interesting philosophical detour—and it’s there for a reason.

Humans have long believed that reasoning is what makes us special.

Historically:

• Ancient philosophers saw reason as humanity’s defining trait.

• Western science and philosophy placed logic and reasoning at the center of knowledge.

• Tools were created to extend human reasoning—math, logic, computers.

AI now threatens to:

• imitate reasoning,

• and possibly redefine it.

That’s unsettling.

The fear isn’t just about job loss.
It’s about loss of uniqueness.

---

The Illusion of Thinking (A Critical Reality Check)

One of the most important sections of the chapter discusses Apple’s 2025 research paper, often referred to as “The Illusion of Thinking.”

The findings are humbling.

Despite impressive performance:

• modern AI models do not truly reason,

• they recognize patterns,

• they imitate reasoning steps,

• but they don’t understand logic the way humans do.

When tasks become:

• deeply logical,

• novel,

• or complex,

AI systems often collapse.

This is called a reasoning collapse.

---

Why This Matters for Agentic AI

Agentic AI systems are usually built on top of large language models.

So these limitations matter.

The chapter emphasizes an important distinction:

Reasoning behavior ≠ reasoning ability

When AI explains its steps, it may look like thinking—but it’s replaying patterns, not understanding cause and effect.

This means:

• autonomy must be constrained,

• self-checking is critical,

• evaluation must be rigorous.

---

Technical and Operational Challenges

Even if reasoning improves, Agentic AI faces serious real-world challenges:

• complex system architecture,

• multi-agent orchestration,

• infrastructure costs,

• reliability and accuracy,

• interoperability with existing systems.

Without solving these, “autonomous AI” risks becoming:

a fragile chain of specialized tools rather than a truly independent system.

---

AI Agents vs Agentic AI: Clearing the Confusion

The chapter spends significant time clarifying a common misunderstanding.

AI Agents

AI agents are:

• software entities,

• designed for specific tasks,

• operating within narrow boundaries.

Examples:

• chatbots,

• recommendation engines,

• robotic vacuum cleaners,

• game-playing bots.

They have autonomy—but limited autonomy.

---

Agentic AI Systems

Agentic AI systems:

• manage complex, multi-step goals,

• coordinate multiple agents and tools,

• adapt workflows dynamically,

• operate over long periods.

They don’t just do tasks.
They manage processes.

---

Why the Distinction Matters

Using these terms interchangeably creates confusion.

Most systems today are:

• AI agents,

• not fully Agentic AI systems.

Understanding the difference helps set:

• realistic expectations,

• proper safeguards,

• appropriate use cases.

---

Strengths and Weaknesses Compared

AI Agents

Strengths

• fast,

• cheap,

• reliable for narrow tasks.

Weaknesses

• brittle,

• limited reasoning,

• poor generalization.

---

Agentic AI Systems

Strengths

• adaptable,

• scalable,

• capable of handling complex workflows.

Weaknesses

• expensive,

• complex,

• still experimental,

• reasoning limitations inherited from models.

---

Prompting Is Not Going Away (At All)

A key message of this chapter is almost counterintuitive:

The more autonomous AI becomes, the more important prompting skills remain.

Why?

Because:

• prompts are how humans express intent,

• prompts are how agents communicate internally,

• prompts act as control mechanisms.

Even inside advanced systems:

• agents pass instructions via prompts,

• reasoning chains are prompt-based,

• coordination often happens through structured language.

---

Prompt Engineering as a Core Skill

Prompting is compared to:

• giving instructions to a smart assistant,

• designing workflows,

• scripting behavior.

It’s not about clever wording.
It’s about clear thinking.

As systems grow more autonomous:

• prompts become higher-level,

• more abstract,

• more strategic.

Prompt engineering evolves into AI system design.

---

Prompting as Control and Safety

Effective prompting can:

• reduce hallucinations,

• constrain unsafe behavior,

• debug agent failures.

In enterprises, prompt libraries are becoming:

• reusable assets,

• cheaper than retraining models,

• critical to quality assurance.

---

The Birth of the Agentic AI Web

One of the most forward-looking sections of the chapter discusses the Agentic AI Web.

The current internet:

• waits for humans,

• reacts to clicks and searches.

The future internet:

• will be navigated by AI agents,

• acting on your behalf,

• behind the scenes.

Instead of visiting websites:

• your agent will talk to other agents,

• compare options,

• complete tasks.

You remain in charge—but you’re no longer doing the busywork.

---

Scaling Agentic AI Beyond Individuals

The chapter goes even further.

Agentic AI could:

• manage cities,

• optimize energy grids,

• coordinate disaster response,

• accelerate scientific discovery.

This requires:

• shared communication standards (like MCP),

• secure data exchange,

• trust-enhancing technologies.

These pieces are emerging—but not fully mature yet.

---

The Shift from E-Commerce to A-Commerce

This is one of the most practical and disruptive ideas in the chapter.

What Is A-Commerce?

A-commerce (autonomous commerce) means:

• AI agents search,

• compare,

• negotiate,

• and purchase on your behalf.

Humans express intent.
Agents handle execution.

---

Why This Changes Everything

Traditional SEO:

• targets human attention.

A-commerce SEO:

• targets AI decision-making.

Websites must become:

• machine-readable,

• structured,

• trustworthy to agents.

If AI agents stop clicking links:

• traffic drops,

• business models change,

• entire industries adapt or disappear.

---

The Final Big Picture

Chapter 1 closes with a powerful insight:

Agentic AI is not about replacing humans.
It’s about changing where humans spend their attention.

Instead of:

• clicking,

• searching,

• comparing,

humans:

• supervise,

• decide,

• set goals.

Children may grow up learning:

• how to manage agents,

• not how to browse the web.

---

Final Takeaway

This chapter sets the stage for everything that follows.

It teaches us that:

• Agentic AI is about autonomy and action,

• reasoning is limited but evolving,

• prompting remains foundational,

• the internet and commerce are changing,

• and responsibility matters as much as capability.

Above all, it reminds us:

The future of AI is not just technical.
It is social, economic, and deeply human.