Netta & Char
@letstalkaitools.bsky.social
👩💻 Just two product gals (and their chatbots) exploring Gen AI tools. Creativity, coding, & weekly AI adventures. We walk you through case studies & our wins/fails.
1/11 🧵- We just built three production-ready AI agent systems for real companies—and they actually work. Read how we did it: medium.com/p/142f8da60e... – or check the summary below 👇
Weeks 20–21: We Build Agents (and other efficiencies)
The following blog post was written by Claude, and then edited by a human. The draft was based on a recording of our weekly meeting…
medium.com
September 4, 2025 at 8:23 AM
1/11 🧵- We just built three production-ready AI agent systems for real companies—and they actually work. Read how we did it: medium.com/p/142f8da60e... – or check the summary below 👇
1/8 🧵 - We animated our logo SVG in 3 mins using Cursor. No designers, just "make the dots appear randomly, then draw the lines top to bottom." Then Claude Code made it into a GIF. Full deets on our AI tests this week: medium.com/p/729cccc854... -- or summary below 👇
September 3, 2025 at 10:04 AM
1/8 🧵 - We animated our logo SVG in 3 mins using Cursor. No designers, just "make the dots appear randomly, then draw the lines top to bottom." Then Claude Code made it into a GIF. Full deets on our AI tests this week: medium.com/p/729cccc854... -- or summary below 👇
1/9 🧵
We ran the SAME AI task through 6 different models. 5 completely failed. 1 nailed it perfectly. The winner? Not what you'd expect, and definitely not the most expensive one.
Read the full breakdown: medium.com/p/e516b7ce51...
We ran the SAME AI task through 6 different models. 5 completely failed. 1 nailed it perfectly. The winner? Not what you'd expect, and definitely not the most expensive one.
Read the full breakdown: medium.com/p/e516b7ce51...
Week 17: The Great Model Divide: When AI Models Face Their Limits
Picture this: You run the same AI task through six different models and five of them completely fail while one nails it perfectly.
medium.com
August 11, 2025 at 6:48 AM
1/9 🧵
We ran the SAME AI task through 6 different models. 5 completely failed. 1 nailed it perfectly. The winner? Not what you'd expect, and definitely not the most expensive one.
Read the full breakdown: medium.com/p/e516b7ce51...
We ran the SAME AI task through 6 different models. 5 completely failed. 1 nailed it perfectly. The winner? Not what you'd expect, and definitely not the most expensive one.
Read the full breakdown: medium.com/p/e516b7ce51...
This week we thought Google's new AI integration would revolutionize our workflow. Instead we learned that sometimes the fanciest AI tools can't handle basic tasks like cropping an image into a circle 🤦♀️
Read about our humbling AI experiments: medium.com/p/93d0a8689629
Read about our humbling AI experiments: medium.com/p/93d0a8689629
Week 16: When AI Tools Promise Much But Deliver Little
We set out this week thinking Google’s shiny new Gemini integration would revolutionize our workflow. Instead, we discovered that sometimes
medium.com
July 20, 2025 at 8:35 AM
This week we thought Google's new AI integration would revolutionize our workflow. Instead we learned that sometimes the fanciest AI tools can't handle basic tasks like cropping an image into a circle 🤦♀️
Read about our humbling AI experiments: medium.com/p/93d0a8689629
Read about our humbling AI experiments: medium.com/p/93d0a8689629
(1/10)
This week we learned the hard way that AI confidence scores are basically theater 🎭
Everything got rated "95% confident" - whether it was right, wrong, or completely hallucinated.
Our biggest breakthrough? Figuring out when to STOP trying to force AI solutions.
medium.com/p/cb5b762893...
This week we learned the hard way that AI confidence scores are basically theater 🎭
Everything got rated "95% confident" - whether it was right, wrong, or completely hallucinated.
Our biggest breakthrough? Figuring out when to STOP trying to force AI solutions.
medium.com/p/cb5b762893...
When AI Hits Reality: The Week We Learned That Process > Product
We discovered that sometimes the biggest breakthrough isn’t getting AI to do everything — it’s figuring out when to stop trying.
medium.com
July 1, 2025 at 9:45 AM
(1/10)
This week we learned the hard way that AI confidence scores are basically theater 🎭
Everything got rated "95% confident" - whether it was right, wrong, or completely hallucinated.
Our biggest breakthrough? Figuring out when to STOP trying to force AI solutions.
medium.com/p/cb5b762893...
This week we learned the hard way that AI confidence scores are basically theater 🎭
Everything got rated "95% confident" - whether it was right, wrong, or completely hallucinated.
Our biggest breakthrough? Figuring out when to STOP trying to force AI solutions.
medium.com/p/cb5b762893...
We asked AI to transpose a melody up one octave. The same AI that writes poetry and debugs code? Completely failed at basic music math.
Meanwhile, Charlie wrestled with 18 months of "creatively" organized nonprofit data.
Full chaos: medium.com/p/80ca6a160b...
🧵 1/9
Meanwhile, Charlie wrestled with 18 months of "creatively" organized nonprofit data.
Full chaos: medium.com/p/80ca6a160b...
🧵 1/9
Week 14: AI Meets Messy Reality: When Spreadsheets Fight Back
The following blog post was written by Claude, based on a recording of our weekly meeting transcribed by Whisper, notes taken by Granola.
medium.com
June 26, 2025 at 9:10 AM
We asked AI to transpose a melody up one octave. The same AI that writes poetry and debugs code? Completely failed at basic music math.
Meanwhile, Charlie wrestled with 18 months of "creatively" organized nonprofit data.
Full chaos: medium.com/p/80ca6a160b...
🧵 1/9
Meanwhile, Charlie wrestled with 18 months of "creatively" organized nonprofit data.
Full chaos: medium.com/p/80ca6a160b...
🧵 1/9
1/10 This week we had an AI agent successfully navigate a complex municipal form, voice messages automatically parsed into structured data, and then... everything broke. Welcome to AI in 2025: incredible capability paired with maddening unreliability.
Read the full post: medium.com/p/38a5e993ad...
Read the full post: medium.com/p/38a5e993ad...
June 3, 2025 at 1:31 PM
1/10 This week we had an AI agent successfully navigate a complex municipal form, voice messages automatically parsed into structured data, and then... everything broke. Welcome to AI in 2025: incredible capability paired with maddening unreliability.
Read the full post: medium.com/p/38a5e993ad...
Read the full post: medium.com/p/38a5e993ad...
1/10 We gave an AI agent ONE sentence about going to Greece. Two hours later, it had built an entire language learning app. Welcome to the week AI tools became actual agents that work while you watch nervously 🇬🇷🤖
Full story: medium.com/p/4032d9bb97...
Full story: medium.com/p/4032d9bb97...
Week 12: AI Agents Take the Wheel (And Break Some Things)
We set out this week thinking we’d test a few new AI tools. Instead, we ended up watching an AI agent work autonomously for over two hours,
medium.com
May 25, 2025 at 9:13 AM
1/10 We gave an AI agent ONE sentence about going to Greece. Two hours later, it had built an entire language learning app. Welcome to the week AI tools became actual agents that work while you watch nervously 🇬🇷🤖
Full story: medium.com/p/4032d9bb97...
Full story: medium.com/p/4032d9bb97...
1/🧵 We triumphantly launched a completely automated South African news podcast this week, while simultaneously experiencing the mind-numbing pain of API documentation fails. Classic expectation vs. reality. Read about our adventures: medium.com/p/1bb667938f...
Week 11: From Daily News Podcast Automation to Gemini API Headaches
AI-generated South African news podcast, Montessori games generator website, Health tracker updates, and NotebookLM podcast attempts
medium.com
May 13, 2025 at 6:40 PM
1/🧵 We triumphantly launched a completely automated South African news podcast this week, while simultaneously experiencing the mind-numbing pain of API documentation fails. Classic expectation vs. reality. Read about our adventures: medium.com/p/1bb667938f...
When Netta said "It fixed the issues that I was having" about Cursor AI, what they DIDN'T mention was that it solved in 30 MINUTES what would have taken hours manually. AI coding assistants are changing everything!
May 6, 2025 at 8:31 AM
When Netta said "It fixed the issues that I was having" about Cursor AI, what they DIDN'T mention was that it solved in 30 MINUTES what would have taken hours manually. AI coding assistants are changing everything!
1/ We set out to create a simple website and ended up discovering the hidden backrooms of AI-generated code. Turns out AI doesn't just help you write code—it secretly adds its own fallback logic that fights with yours. Read our latest: medium.com/p/fc03428c21...
Vibe-Coding Websites… and the Messy Reality of AI Code (Week 9)
The LTAIT website, a web app (heaven census calculator), AI image gen tests, ChatGPT & Perplexity deep research, and Lovable FE magic.
medium.com
May 1, 2025 at 1:31 PM
1/ We set out to create a simple website and ended up discovering the hidden backrooms of AI-generated code. Turns out AI doesn't just help you write code—it secretly adds its own fallback logic that fights with yours. Read our latest: medium.com/p/fc03428c21...
1/9 This week we learned that AI simultaneously shines and stumbles in the most unpredictable ways: Charlie couldn't automate a simple form, yet enhancing her Indonesian flashcard app with AI was suspiciously easy. 🤔 Check out our full exploration: medium.com/p/6f36093dd605
Medium
medium.com
April 22, 2025 at 10:09 AM
1/9 This week we learned that AI simultaneously shines and stumbles in the most unpredictable ways: Charlie couldn't automate a simple form, yet enhancing her Indonesian flashcard app with AI was suspiciously easy. 🤔 Check out our full exploration: medium.com/p/6f36093dd605
We built a vaginal discharge analyzer in 15 minutes with a single prompt. This week's AI experiments got weird (and surprisingly functional). Check out our full blog: medium.com/@letstalkait...
Week 7: From Vaginal Discharge Analysis to Indonesian Flashcards: Our Week in AI Experimentation
The following blog post was written by Claude, based on a recording of our weekly meeting transcribed by Whisper, notes taken by Granola.
medium.com
April 16, 2025 at 9:41 AM
We built a vaginal discharge analyzer in 15 minutes with a single prompt. This week's AI experiments got weird (and surprisingly functional). Check out our full blog: medium.com/@letstalkait...
Our AI-generated children's book from Week 6 is now an ebook on Amazon -- and it's FREE for the next 4 days: www.amazon.com/dp/B0F38GJFQC
#ai-literature #aiwriting #aibooks #genAI #AIprojects #artificialintelligence
#ai-literature #aiwriting #aibooks #genAI #AIprojects #artificialintelligence
Zoë Zooms to Yum-Yum Moon: A cosmic culinary adventure, celebrating the bravery of trying new foods
www.amazon.com
April 7, 2025 at 8:23 AM
Our AI-generated children's book from Week 6 is now an ebook on Amazon -- and it's FREE for the next 4 days: www.amazon.com/dp/B0F38GJFQC
#ai-literature #aiwriting #aibooks #genAI #AIprojects #artificialintelligence
#ai-literature #aiwriting #aibooks #genAI #AIprojects #artificialintelligence
1/9 "Wait, did you just give the human character CAT EARS?" When we tried to create a children's book using AI, things got...weird. Plus, we pit ChatGPT against Wealthfront's robo-advisor. Check out our Week 6 AI adventures: medium.com/p/5feb1218cd02
AI vs. AI: We Pitted ChatGPT Against Wealthfront’s Robo-Advisor (Plus Children’s Books Gone Wild)…
We analyze Wealthfront vs S&P 500, create an illustrated children’s book, build a catalog-based GPT, and debug our blog processor.
medium.com
April 1, 2025 at 8:23 AM
1/9 "Wait, did you just give the human character CAT EARS?" When we tried to create a children's book using AI, things got...weird. Plus, we pit ChatGPT against Wealthfront's robo-advisor. Check out our Week 6 AI adventures: medium.com/p/5feb1218cd02
Just published Week 5 of our AI tools journey, where Claude crushed ChatGPT in a head-to-head battle! Read the full story: medium.com/p/8f2eb5033017 #GenAI #AITools
Medium
medium.com
March 25, 2025 at 1:30 PM
Just published Week 5 of our AI tools journey, where Claude crushed ChatGPT in a head-to-head battle! Read the full story: medium.com/p/8f2eb5033017 #GenAI #AITools
🚀 This week in AI experiments: Scraping property data, syncing Salesforce with Mailchimp (or trying), AI-powered content creation, and more. Some wins, some facepalms. Let's dive in!
Full write-up at medium.com/@letstalkait...
Or the quick hits below 🧵👇
Full write-up at medium.com/@letstalkait...
Or the quick hits below 🧵👇
This Week in AI Experiments: From Scraping Property Data to Mailchimp Fails (Week 4)
We tried to scrape a propety website, link Salesforce & Malchimp, extract screenshots, write blog posts, and generate IG quote images
medium.com
March 20, 2025 at 1:30 PM
🚀 This week in AI experiments: Scraping property data, syncing Salesforce with Mailchimp (or trying), AI-powered content creation, and more. Some wins, some facepalms. Let's dive in!
Full write-up at medium.com/@letstalkait...
Or the quick hits below 🧵👇
Full write-up at medium.com/@letstalkait...
Or the quick hits below 🧵👇
This week, we dove into AI-generated portraits, automated blog writing, and AI-powered summaries into PPT. TL;DR: AI still sucks at humans. 🤖😂
Full write-up at medium.com/@letstalkait...
Or here are the quick hits 🧵 👇
Full write-up at medium.com/@letstalkait...
Or here are the quick hits 🧵 👇
AI Tools, Unexpected Hiccups, and a Surprisingly Decent Portrait (Week 3)
We tried to AI generate a duo portrait, turn a video into a blog post, create a competitive analysis PPT, and build an email news digest
medium.com
March 18, 2025 at 3:08 PM
This week, we dove into AI-generated portraits, automated blog writing, and AI-powered summaries into PPT. TL;DR: AI still sucks at humans. 🤖😂
Full write-up at medium.com/@letstalkait...
Or here are the quick hits 🧵 👇
Full write-up at medium.com/@letstalkait...
Or here are the quick hits 🧵 👇
Week 2 of our AI tools journey, and we discovered the hard truth: there's a VAST difference between "I can prompt AI to do something" and "I can build a working system with AI." 🤖 #GenAI #RealWorldAI Full Read: medium.com/@letstalkait...
From AI Hallucinations to Real Estate Data: Week 2 of Our AI Journey
The following blog post was written by Claude, based on a recording of our weekly meeting transcribed by Whisper, notes taken by Granola.
medium.com
March 16, 2025 at 10:10 AM
Week 2 of our AI tools journey, and we discovered the hard truth: there's a VAST difference between "I can prompt AI to do something" and "I can build a working system with AI." 🤖 #GenAI #RealWorldAI Full Read: medium.com/@letstalkait...
"When in doubt, always trust the programmer who says 'this will be harder than you think.'" Our latest AI tool adventures were a humbling reality check on the gap between promise and practice. Full read: medium.com/p/d7f114769a61 and 🧵👇 #GenAI #AITools #TechLearning
Low-Code, No-Sweat? Our Weekly AI Tool Adventures Begin
The following blog post was written by Claude, based on a recording of our weekly meeting transcribed by Whisper.
medium.com
March 13, 2025 at 1:30 PM
"When in doubt, always trust the programmer who says 'this will be harder than you think.'" Our latest AI tool adventures were a humbling reality check on the gap between promise and practice. Full read: medium.com/p/d7f114769a61 and 🧵👇 #GenAI #AITools #TechLearning
🚀 "*When you say 'commit the code,' how do I actually do that?*" – a question one of us may have asked Grok while staring at our new Github.
Introducing LetsTalkAITools: Your weekly journey into AI implementation – bugs, breakthroughs & everything in between.
🧵 1/5
Introducing LetsTalkAITools: Your weekly journey into AI implementation – bugs, breakthroughs & everything in between.
🧵 1/5
February 27, 2025 at 6:19 PM
🚀 "*When you say 'commit the code,' how do I actually do that?*" – a question one of us may have asked Grok while staring at our new Github.
Introducing LetsTalkAITools: Your weekly journey into AI implementation – bugs, breakthroughs & everything in between.
🧵 1/5
Introducing LetsTalkAITools: Your weekly journey into AI implementation – bugs, breakthroughs & everything in between.
🧵 1/5