- Product Upfront AI
- Posts
- Who's using AI? Claude's conversations revealed the truth about AI at work
Who's using AI? Claude's conversations revealed the truth about AI at work
How to build in public without sounding like everyone else

"Everyone's talking about AI replacing jobs!"
But have you seen what's actually happening in the workplace? (Not the doomsday headlines, but the real data?)
We all imagine robots taking over offices.
But here's the reality: The latest Anthropic research just blew that myth wide open.
They discovered AI isn't replacing us - it's becoming our digital sidekick for:
Writing code smarter
Creating better documentation
Getting work done faster
We worry about total automation... while missing the real story.

More findings from the Economic Index’s first paper
AI isn't stealing jobs. It's transforming HOW we work - and mostly in higher-paid roles.
Here's the fascinating part: Only 4% of jobs use AI heavily, but over a third are starting to embrace it.
And guess what?
It's not about replacement - 57% of the time, we're working alongside AI, not being replaced by it.
Today, I'm sharing where the real opportunities are in this AI revolution (and they're not what the headlines would have you believe).
"The REAL Truth About AI at Work!"
Ready for some mind-blowing insights about how AI is actually changing our jobs? (Not the hyped-up predictions, but real data!)
The Surprising Reality:
Creative Tasks: Think designing, writing, and editing - they're seeing massive AI adoption
Visual Pattern Tasks: From security screening to radiology, AI is changing how we spot details
Task-Specific Impact: Different jobs share similar tasks that AI excels at (and that's where the magic happens)
But here's what's really interesting...
AI isn't replacing entire jobs - it's transforming specific tasks across multiple professions.

We're seeing AI adoption in:
Mid-Level Tasks: Tasks that blend human expertise with AI assistance
Cross-Professional Skills: Similar tasks appearing in completely different careers
Collaborative Work: Where AI assists rather than replaces
And the biggest revelation?
The impact isn't about whole jobs disappearing - it's about how tasks are being enhanced across different roles.
Drop a 🤔 if you're seeing these changes in your field!
Want to know what's exciting about all this?
(Hint: We're just scratching the surface)
Here's what's about to unfold, and trust me, it's mind-blowing:
AI capabilities are exploding
The job landscape is shifting faster than ever
We're tracking every move with the Anthropic Economic Index
But here's the game-changing part...
We're about to get a front-row seat to see:
Which jobs evolve vs. which ones vanish
Where AI becomes your co-pilot vs. your replacement
How deep AI goes into different careers
💡 The Million-Dollar Question:
Will we keep seeing AI handle specific tasks while leaving most jobs intact? (The answer could reshape your entire career path)
"How The Associated Press Actually Uses AI in News Production"
Let's look at a real case that perfectly connects to our newsletter's theme about AI enhancing rather than replacing jobs:
The Associated Press (AP), one of the world's largest news organizations, provides a perfect real-world example of what we discussed about AI transforming specific tasks rather than entire jobs.
Here's Their Actual AI Implementation: AP uses AI tools to handle data-heavy stories, particularly in:
Financial earnings reports
Sports recaps
Market analysis
The Real Numbers (Publicly Shared by AP):
Automated writing now handles 3,700 earnings reports per quarter
This is 12x more than their reporters could cover manually
Key Point: Not a single journalist was replaced
The Actual Workflow:
AI handles the initial data analysis and basic story structure
Human journalists then:
Add context and insights
Interview sources
Investigate deeper angles
Write more complex analysis pieces
Concrete Results:
Journalists report spending 20% more time on investigative stories
Coverage of smaller companies increased by 300%
Accuracy in financial figures improved significantly
This perfectly illustrates our newsletter's point about AI becoming a "digital sidekick" - AP's journalists aren't being replaced; they're being freed up to do more impactful work.
Source: This information comes from AP's own public statements and their partnership documentation with Automated Insights.
Why This Matters:
This shows exactly what we discussed earlier - AI handling specific tasks (data processing, basic writing) while humans focus on higher-value work (analysis, relationships, complex storytelling).
Want to learn more about AP's approach?
They've published their AI standards and practices guide online, which provides excellent insights for any media professional looking to integrate AI responsibly.
This is just one real example of what we discussed in our newsletter
The Complete Guide: SaaS Founder's AI Twitter Arsenal
(The detailed prompts I actually use every day)
Let's transform those basic prompts into power templates that generate scroll-stopping content:
1. The Thread Architect™ (Extended Version)
Create a Twitter thread about [specific SaaS topic] following these guidelines:
HOOK TWEET:
- Start with a pattern interrupt ("Everyone's wrong about [X]" or "Want to know why [uncommon observation]?")
- Include one surprising stat or counterintuitive statement
- End with curiosity gap
- Max 280 characters
BODY (5-7 tweets):
- Each tweet should be standalone valuable
- Include at least 2 tweets with specific numbers/data
- Add 1 tweet with personal experience/story
- Include 1 "comum mistake to avoid"
- Format: Problem → Solution → Result
CLOSING TWEET:
- Actionable summary
- Clear CTA
- Hint at additional value
Style notes:
- Use emojis strategically (max 2 per tweet)
- Include spacing for readability
- Add brackets [like this] for emphasis
- Use * for highlighting key points
2. The Metric Storyteller (Detailed Version)
Help me create a compelling tweet about [specific metric/milestone] using this framework:
SETUP:
Context: [Brief background]
Metric: [Specific number/achievement]
Time period: [Timeframe]
FORMAT:
1. Opening line: Use contrast ("We went from X to Y" or "Everyone said X, but we found Y")
2. Share the unexpected challenge
3. Key learning moment
4. Specific numbers (with comparison)
5. Tactical insight others can use
MUST INCLUDE:
- One specific mistake that led to learning
- Actual numbers (not percentages only)
- One unexpected insight
- Something that didn't work
TONE:
- Humble but confident
- Data-driven but relatable
- Teaching > Bragging
Max length: 2-3 tweets
3. The Customer Win Spotlight (Advanced)
Create a customer success story tweet sequence:
STRUCTURE:
Tweet 1 - The Setup:
- Industry-specific pain point
- Unexpected twist
- Hint at result
Tweet 2 - The Journey:
- Specific challenge faced
- Common solution that failed
- Our unique approach
Tweet 3 - The Results:
- Lead with specific metrics
- Include unexpected benefit
- Customer quote (if available)
REQUIREMENTS:
- Focus on customer's journey, not our product
- Include at least 2 specific metrics
- Add one relatable struggle
- Mention timeline
- Include learning that applies to others
AVOID:
- Generic success claims
- Product feature lists
- Technical jargon
4. The Feature Launch Formula (Expanded)
Create a feature announcement that converts:
PRE-LAUNCH TWEET:
- Tease problem being solved
- Hint at unique solution
- Create FOMO
LAUNCH TWEET:
1. Problem statement (relatable scenario)
2. Unexpected solution angle
3. Key benefit (in numbers if possible)
4. Social proof element
5. Clear next step
FOLLOW-UP:
- One unexpected use case
- Early user quote
- Limited time offer (if applicable)
STYLE GUIDELINES:
- Use power words sparingly
- Include whitespace
- Add (1/3) format for thread
- End with clear CTA
BONUS: Add engagement hook:
"Reply with __ if you want early access"
5. The Founder Journey Share (Deep Dive)
Create a vulnerable yet valuable founder story:
FOUNDATION:
Main learning: [Key insight]
Timeline: [When it happened]
Result: [Specific outcome]
NARRATIVE STRUCTURE:
1. Hook: Common belief you had wrong
2. Context: What led to the situation
3. Crisis moment: Specific challenge
4. Learning: Key insight that changed everything
5. Results: Tangible outcomes
6. Lesson: Actionable takeaway
KEY ELEMENTS TO INCLUDE:
- One specific mistake
- Actual numbers/metrics
- Time it took to learn this
- What you'd do differently
- Resource that helped
FORMATTING:
- Break into 3-4 tweets
- Use → for cause/effect
- Include timestamps
- Add parentheses for asides
- Use line breaks for emphasis
VOICE:
Honest > Perfect
Specific > Generic
Story > Advice
💡 Pro Tips for All Templates:
Test hooks with different angles
Save best-performing variants
Track which CTAs convert best
Always include one unexpected insight
Use timestamps for credibility
P.S. Next week I'm sharing more on this…SO STAY TUNE!
Before We Go...
Thanks for diving deep with me today!
If you found these prompts helpful, hit that like button and share it with another founder who's trying to build in public.
Next week, we're going even deeper with prompts that helped SaaS founders:
Triple their content output
Generate viral threads consistently
Build engaged communities
See you next Wednesday!
Reply