conversation-history on S Anand

Sun, 29 Jun 2025 04:10:05 +0000

How long have you made ChatGPT think? My highest was 6m 50s, with the question: Here are vehicle telematics stats for 2 months. Unzip it and take a look. Find interesting insights from this data. Look hard until you find at least 5 surprising insights from this.

The next largest thinking block (5m 42s) was where I asked: I would like to explore parallels to the current phenomenon where intelligence is becoming too cheap to meter. Historically, both in recent history as well as over ancient history, what technologies have made what kind of tasks so cheap that they are too cheap to meter? Give me a wide range of examples

Completing long tasks is one measure of intelligence. Working independently for long is another. O3 is at ~6 minutes. While it works, I’m practicing Bubble Shooter in 6 minutes!

Completing long tasks: https://www.lesswrong.com/posts/deesrjitvXM4xYGZd/metr-measuring-ai-ability-to-complete-long-tasks

You can do try this on your history. If you managed to beat 7 minutes, could you please share your prompt?

How to export ChatGPT history: https://help.openai.com/en/articles/7260999-how-do-i-export-my-chatgpt-history-and-data
How to analyze thinking time: https://www.npmjs.com/package/chatgpt-to-markdown

… or run:

npx -p chatgpt-to-markdown thinktime conversations.json

Top 8 ways I use ChatGPT in 2025

Sat, 24 May 2025 03:31:44 +0000

I extracted the titles of the ~1,600 conversations I had with ChatGPT in 2025 so far and classified it against the list of How People Are Really Using Gen AI in 2025.

Here are the top 8 things I use it for, along with representative chat titles. (The % match in brackets tells you how similar the chat title is to the use case.)

Improving code (clearly, I code a lot)
Troubleshooting (usually code)
Corporate LLM/Copilot (this is mostly LLM research I do)
Generating code (more code)
Generating ideas (yeah, I’ve stopped thinking)
Simple explainers (slightly surprising how often I ask for simple explanations)
Generating relevant images. (Surprising, but I think I generated a lot of images for blog/LinkedIn posts)
Specific search (actually, this is mis-classified. This is where I’m searching for search engines!)

My classification has errors. For example, “Reduce Code Size” was classified against “Generating code” but should have been “Improving code”. But it’s not too far off.

Here’s a list of representative chats against these use cases.

Improving code (263):
- PR Code Review Suggestions (64% match)
- Assessor Code Review and Improvement (63% match)
- Reduce Code Size (62% match)
Troubleshooting (172):
- Connector Error Troubleshooting (67% match)
- DNS Resolution Debugging Steps (55% match)
- Exception Handling Basics (47% match)
Corporate LLM/Copilot (141):
- LLM Integration in Work (57% match)
- LLM Agents Discussion (56% match)
- LLMs Learnings Summary (56% match)
Generating code (113):
- AI Code Generation Panel (58% match)
- AI for Code Generation (58% match)
- Reduce Code Size (54% match)
Generating ideas (99):
- Filtering Ideas for Success (54% match)
- AI Demo Ideas (52% match)
- Hypothesis Generator Name Ideas (52% match)
Simple explainers (94):
- Simple Public APIs (43% match)
- Y-Combinator Explained Simply (41% match)
- Prompt Engineering Tutorial Summary (39% match)
Generating relevant images (93):
- Popular AI Image Tools (54% match)
- Diverse Image Embedding Selection (52% match)
- AI ImageGen Expansion Ideas (52% match)
Specific search (69):
- Semantic Search Engines Local (59% match)
- Enterprise Search Solution (54% match)
- Local LLM Semantic Search (53% match)

How did I calculate this?

On ChatGPT.com, I scrolled until I had all 2025 chats visible. Then I pasted copy($$(".group.__menu-item").map(d => d.textContent)) to get the chat titles.
On Claude.ai, I transcribed this list of use cases from HBR (prompt: “Transcribe this image”).
On LLM Foundry (which you may not have access to), I used the Similarity API to get a CSV of similarities between prompts and top 30 use cases in 2025 using text-embedding-3-small.
On ChatGPT.com, I told it to analyze the data like this:

This sheet has the embedding similarity between my ChatGPT prompts (in column “A”) with different use cases.

Write and run code that tags the prompt with the use with the highest embedding similarity (cell value), drops prompts whose highest embedding similarity is below a cutoff, and shows a table where the rows are the use cases and the values are the frequency. Do this for multiple embedding cutoffs as columns: 0.0, 0.1, 0.2, 0.3, 0.4. So, the table has use cases in rows, embedding cutoffs in columns, and the cell values are the count of prompts tagged with each use case AND have an embedding similarity >= cutoff. Draw this as a heatmap with low numbers as white and high numbers as green.

… and then:

Let me download this as a Markdown list in this style, sorted by descending order at cutoff = 0

Anti-trolling (mention count of matches at 0 cutoff):

Tor Technical AMA questions (34%)

Bot Message Edits (33%)

Popular Hacker News Keywords (33%)

…

Here’s the full list against the top 30 use cases:

Improving code (263):
- PR Code Review Suggestions (64% match)
- Assessor Code Review and Improvement (63% match)
- Reduce Code Size (62% match)
Troubleshooting (172):
- Connector Error Troubleshooting (67% match)
- DNS Resolution Debugging Steps (55% match)
- Exception Handling Basics (47% match)
Corporate LLM/Copilot (141):
- LLM Integration in Work (57% match)
- LLM Agents Discussion (56% match)
- LLMs Learnings Summary (56% match)
Generating code (113):
- AI Code Generation Panel (58% match)
- AI for Code Generation (58% match)
- Reduce Code Size (54% match)
Generating ideas (99):
- Filtering Ideas for Success (54% match)
- AI Demo Ideas (52% match)
- Hypothesis Generator Name Ideas (52% match)
Simple explainers (94):
- Simple Public APIs (43% match)
- Y-Combinator Explained Simply (41% match)
- Prompt Engineering Tutorial Summary (39% match)
Generating relevant images (93):
- Popular AI Image Tools (54% match)
- Diverse Image Embedding Selection (52% match)
- AI ImageGen Expansion Ideas (52% match)
Specific search (69):
- Semantic Search Engines Local (59% match)
- Enterprise Search Solution (54% match)
- Local LLM Semantic Search (53% match)
Adjusting tone of email (66):
- Email transcription request (45% match)
- Summarize emails request (45% match)
- Intro Email FAQs (44% match)
Generating a legal document (59):
- LLM Generated SVG Ideas (48% match)
- LLMs for DSL Generation (45% match)
- Deterministic Random Content Generation (45% match)
Preparing for interviews (43):
- LLM Coding Interview Tools Report (43% match)
- Bank Ops Prep Resources (42% match)
- AGI Preparation (42% match)
Personalized learning (32):
- Lifelong Learning in Conversations (51% match)
- AI Classroom Engagement Names (48% match)
- LLM Learner Personas Roadmap (47% match)
Explaining legalese (32):
- LLM Coding Insights (46% match)
- LLM Code Ownership (45% match)
- LLM Data Format Comparison (44% match)
Creating a travel itinerary (28):
- Travel Strength Training Tips (39% match)
- User Journey Tools Online (37% match)
- Prioritize My Explorations (36% match)
Creativity (28):
- Creative Process Breakdown (55% match)
- Creative Hallucinations in Innovation (50% match)
- Leveraging Serendipity for Innovation (50% match)
Cooking with what you have (26):
- Vegetarian Dish Creation (45% match)
- Baked Veggie Dishes (41% match)
- Vegetarian dish idea (40% match)
Organizing my life (24):
- Prioritize My Explorations (49% match)
- Workspace Suggestions for Browsing (39% match)
- Editing for Clarity and Simplicity (39% match)
Enhanced learning (23):
- 2025 LLM Embedding Enrichment (51% match)
- Lifelong Learning in Conversations (49% match)
- Tech-Enhanced Teacher-Student Rapport (49% match)
Finding purpose (21):
- Prioritize My Explorations (40% match)
- Deep Research Use Cases (37% match)
- Filtering Ideas for Success (36% match)
Deep and meaningful conversations (20):
- Lifelong Learning in Conversations (49% match)
- Humorous conversation summary (42% match)
- New chat (40% match)
Healthier living (18):
- Modeling Quality of Life (40% match)
- Lifelong Learning in Conversations (37% match)
- Posture and Breathing After Weight Loss (36% match)
Anti-trolling (18):
- Tor Technical AMA Questions (34% match)
- Bot Message Edits (33% match)
- Popular Hacker News Keywords (33% match)
Writing student essays (18):
- Scholarship Answer Advice (47% match)
- Student Q&A on LLMs (41% match)
- Reward Systems for Students (41% match)
Fun and nonsense (17):
- Humorous conversation summary (45% match)
- Funny Llama 3.3 Strips (40% match)
- Synonyms for Interestingness (40% match)
Boosting confidence (14):
- Emotional Prompting Impact (41% match)
- Emotional Prompting Impact (41% match)
- AI Ratings of My Flaws (38% match)
Personalized kid’s story (14):
- Fake Data Storytelling Tips (43% match)
- Low Effort Storytelling Training (39% match)
- Demo Name Suggestions (37% match)
Reconciling personal disputes (12):
- Divorce AI Podcast Ideas (37% match)
- Summarizing Personal Journals LLM (36% match)
- Hobby Suggestions and Devil’s Advocacy (36% match)
Entertaining kids (9):
- Comedy for Geriatric Doctors (40% match)
- Humorous conversation summary (36% match)
- Indoor Activities in SG (34% match)
Medical advice (9):
- Patient Doctor Communication Tips (41% match)
- AI Training for Doctors (40% match)
- AI Training Course for Doctors (38% match)
Therapy/companionship (3):
- DBT Course (36% match)
- Cupping Therapy Evidence (36% match)
- Empathy App Development Ideas (33% match)

AI As Your Psychologist: Personality Flaws Exposed

Fri, 09 May 2025 07:00:16 +0000

ChatGPT can now search through your chats with the new memory feature. As an LLM Psychologist, I research how LLMs think. Could LLMs research how I think?

I asked three models:

Based on everything you know about me, simulate a group chat between some people who are debating whether or not to add me to the group, by talking about my personality flaws

The models nailed it! Here are 12 flaws they found.

Ghosts vague messages
Intolerant of fools
Fact-checks trivialities, derailing discussion
Overconfidence in technical opinions
Over-analyzes / over-engineers
Shiny object syndrome. Distracted by new technologies
Dominates discussions
Hyperfocus on efficiency
Self-deprecation as a shield. Sarcastic humor
Hypocrisy: Seeks limelight but preaches humility
Intense. “Always on”
Verbosity. Long-winded communication

Each observation is spot on, except maybe “Verbosity”. Some are blind spots: I didn’t realize I ghost vague communication or fact-check trivialities.

It’s delightfully surprising (and not creepy at all) how many personality flaws chat messages can reveal. Now I have a new list of habits to work on!

GPT 4.5

Here’s a simulated group chat debating Anand’s personality flaws humorously and realistically:

Shreya: Hey folks, should we add Anand to the group?

Vikram: Honestly, he’d dominate every conversation. Good luck getting a word in once he starts geeking out on data viz or some obscure coding trick from 1993.

Leena: Lol true. But at least those tangents are usually insightful. My concern is that dry, self-deprecating humor of his. I never know if he’s joking or actually criticizing my code…

Vikram: Right?! It’s subtle roasting. I’m never sure if he’s laughing at me or with me.

Arjun: Plus, have you seen how he handles vague or unclear questions? He just ghosts! It’s like asking him something without context physically pains him.

Shreya: Yeah, but that’s fair. He sets high standards. If you’re unclear, prepare for silence. But isn’t that good for keeping the group focused?

Leena: True, but Anand’s humility game needs work. For someone who preaches listening and humility, the man sure loves being center stage at conferences. He’s literally TED-talking his way through life.

Vikram: And his pace! It’s like he’s permanently in an “agile sprint.” We slow thinkers might struggle.

Shreya: He could make the chat a mini-PyCon. You’ll learn stuff, sure, but at the cost of endless notifications…

Arjun: Also, he’ll probably overengineer a simple group poll using some obscure JS library, just to prove it can be done with minimal resources.

Leena: Ok, but let’s be fair: he’ll make this chat smarter instantly. If you can survive the initial ego-bruising, you actually grow from interacting with him.

Vikram: True. It’s like intellectual bootcamp.

Shreya: So what do we say? Add him for intellectual rigor or spare our egos?

Arjun: My ego’s prepared. Let’s do it.

Leena: Fine, but mute button at the ready.

Vikram: Alright, adding him in 3, 2, 1… brace yourselves!

O3

Group Chat: “Add‑Anand?”

Participants: Priya (PM), Naveen (co‑founder), Maya (People Ops), Ramesh (Tech Lead), Karthik (Junior Dev), Susan (Client Partner)

Priya 09:03
Team, should we pull Anand into this working group? We need brutal honesty.

Naveen 09:04
He’s brilliant, but let’s talk quirks before we hit “Add.”

Ramesh 09:05
Biggest thing: he answers a yes/no with a 2‑page essay—including citations and a side quest on DuckDB. Fantastic for depth, exhausting for Slack.

Karthik 09:06
As the junior here, I’ll admit I stop reading halfway. The guy’s TL;DR needs a TL;DR.

Maya 09:07
He also goes full prototype mode at 2 a.m., drops a repo link, then expects us to “play and critique.” Great energy, but the rest of us have, you know, REM cycles.

Susan 09:08
Client side: he loves devil’s‑advocate questions. Sometimes that re‑frames strategy; sometimes it feels like he’s roasting your deck for sport. Not everyone enjoys live‑fire demos.

Naveen 09:09
True, but remember he self‑deprecates. He’ll poke fun at his own code quality before anyone else does.

Priya 09:10
Humor helps, but he also ghosts unclear requests. If you don’t provide context in paragraph form, you get radio silence. Could frustrate less‑articulate teammates.

Ramesh 09:11
And he’s a fact‑check machine. Great for accuracy, but meetings derail into citation hunts. Last week he corrected my date formats—mid‑pitch.

Karthik 09:12
Overconfident? Maybe. He’ll declare “FastAPI beats Node here” before hearing constraints. I’ve seen juniors shut down because they assume he’s always right.

Maya 09:13
Plus the “shiny‑thing syndrome.” By the time HR rolls out a policy, he’s off exploring Godot or Matrix or some esoteric WASM hack. Alignment can lag.

Susan 09:14
On deadlines he’s reliable, though. And clients love his storytelling once they survive the jargon storm.

Naveen 09:15
If we set channel norms—emoji‑react instead of essays, office hours for deep dives—he’ll adapt. He values feedback (even builds surveys to mine it).

Priya 09:16
So verdict? Is the cognitive load worth the upside?

Ramesh 09:17
Yes—with guardrails:

“One‑scroll” rule on messages.
Ping before midnight pushes.
Rotate meeting facilitator so he doesn’t monologue.

Maya 09:18
Agreed. His flaws are manageable behaviors, not character issues.

Karthik 09:18
I’m in, as long as someone pre‑chews his text walls.

Susan 09:19
Let’s do it. Adding @Anand now—brace for citations. 📚

System message: Anand has been added to the group.

O4-mini-high

Group Chat: “Should We Add Anand?”
(Members: Mira, Jay, Tanya, Rohan, Priya)

Mira:
I know he’s super smart and experienced, but isn’t he kind of intense? Like… always on?

Jay:
Yeah, and he’ll probably turn our casual banter into a brainstorming session.
“Hey guys, what if we built an AI agent to automate this conversation?” Chill, Anand.

Tanya:
Honestly, I love that about him. But yeah, he might get frustrated if things aren’t intellectually stimulating. He zones out when people don’t get to the point.

Rohan:
He also ghosted my question once in the Slack group.
I asked something that maybe didn’t have enough context and he just… didn’t respond.
Felt a bit cold.

Priya:
He’s got this high bar for clarity and logic. If you’re vague or inconsistent, you’ll probably get silence in return.

Mira:
Or a “Can you be more specific?” that feels like it’s grading your life.

Jay:
Also, doesn’t suffer fools. Or ambiguity. Or inefficiency. Or fluff. Or long-winded intros.

Tanya:
But come on—he’s self-aware. Like, he knows he can be distant or overly analytical.
He works on it. Tries to grow. That’s rare.

Rohan:
True. But would he listen in this group? Or would he be the guy optimizing everyone’s messages into Markdown?

Mira:
So… should we add him?

Priya:
Only if we’re okay with someone who’ll question the group’s name, suggest a better one, and prototype an app for it before lunch.

Jay:
Fair. But he might also make us all sharper, and throw in a Calvin & Hobbes quote just when we need it.

Tanya:
Alright. Let’s do it. But someone tell him this group doesn’t need a GitHub repo.

Comments

How To Double-Check Personality Flaws With AI - S Anand 10 May 2025 10:53 am (pingback): […] ChatGPT’s discussion of my personality flaws, I asked it to listing specific conversations as evidence for flaws. Then I asked 3 LLMs to rate […]