Senior Developer Advocate at Atlassian | Co-Host of The Programming Podcast | Founder of @commityourcode | Prev Google, AutoZone, and A Local Gas Station.
Dallas, TX
Joined January 2019
- Tweets28.4K
- Following2.8K
- Followers301K
- Likes14K
Pinned Tweet
I JUST RECEIVED THE FINAL VERSION OF THE BOOK!🥹
We poured EVERYTHING into this book. Many books focus on AI from the perspective of the highest levels of education but what about the average corporate dev that wants to stay ahead & do well at their jobs?
We wrote this for you!
I had Grok @bot go and remove my info and my wife's info from HUNDREDS of Data brokers. It took me almost no effort at all besides Asking bot to do it.
On an average day I would get tons of spam calls and no matter what I would do they would come in. Grok bot started this on Monday. Today? Zero calls. Not a single one.
My inbox is flooded with emails from the data brokers that have successfully removed our info!
For those asking for the prompt, you can use this
Remove my household's personal information from every data broker and people-search site I can reach, not a sample. The goal is fewer spam calls, texts, and emails. Prioritize brokers that sell data to other brokers, but cover all of them.
PEOPLE (repeat for each person)
- Full legal name, plus nicknames, maiden names, and other spellings: [ ]
- Date of birth: [ ]
- Current address: [ ]
- Previous cities/addresses (last 10–15 years): [ ]
- Phone numbers, and whose each one is: [ ]
- Email to use for requests: [ ] (a separate email per person if possible, since some sites allow only one identity per email)
- State of residence: [ ] (cite that state's privacy and data broker laws)
BUILD THE MASTER LIST
1. Download the complete state data broker registries: California (cppa.ca.gov, downloadable with privacy emails and opt-out links) plus Texas, Vermont, and Oregon where possible. Add every people-search site you can find from current opt-out guides.
2. Remove duplicates, personal Gmail/Yahoo addresses, and marketing agencies. Save it as a spreadsheet on your computer.
3. Sort every entry into one of two groups: web opt-out form, or email-only.
WORK THE WHOLE LIST
4. Web forms: in batches of 10–15, search each site for each person's records and submit delete plus do-not-sell/share requests. Keep going until every form site is done.
5. Email-only: draft a deletion and opt-out request citing my state's law that lists everyone above. Send it in batches of about 40–50 recipients per email until every address has been covered. Show me each email as a draft first; I'll press Send.
6. Confirmations: watch my inbox, click every verification link, and finish any form that a reply points to. Brokers that say "use our web form" move to the form list.
RULES
- No ID uploads, no SSN (not even the last 4), no payments, no password accounts, no phone calls. Skip anything that needs those and list it for me.
- Never reply to a broker or email anyone new without showing me a draft first.
- Ask me only for things only I can do (texted codes, an email address for another person).
TRACK AND FOLLOW UP
7. Keep a tracker with each broker and each person's status: submitted, confirmed, no record, bounced, skipped (and why), or needs me.
8. After each round, give me a running total. Keep going in rounds until the list is empty or I say stop.
9. Set up follow-up checks for tomorrow and again in 3–4 weeks to recheck the people-search listings, click new links, tally replies and bounces, and re-submit anything that came back.
Software Developers keeping making this same mistake with AI agents, your confusing instructions with enforcement.
You create an AGENTS.md file, you tell the agent to run tests, review its code, clean things up, AND follow your engineering standards.
But... what happens when the agent skips a step?
What happens when it confidently tells you everything is working, but nobody actually verified the results?
If your quality gate is "the AI said it checked," you don't have a quality gate. This is where agentic engineering gets REALLY interesting.
Instead of hoping the model follows instructions, you can build workflows that enforce certain actions, capture results, and prevent the next step from happening until requirements are satisfied.
Another agent reviewing the work can help, but that alone isn't the same as deterministic verification.
We got into this on the latest episode of The Programming Podcast, including how tools like Atomic are approaching verifiable agent runtimes.
The real unlock isn't getting AI agents to do more work.
It's engineering systems that can verify the work they're doing.
This is from the latest episode of The Programming Podcast with our guest cohost @RayFernando1337!
Spotify
open.spotify.com/episode/4F6…
Youtube
youtube.com/watch?v=y-Kb3RSk…
Danny Thompson retweeted
Critical thinking and ability to effectively communicate processes is an essential skill requirement that cuts across different disciplines.
One of the biggest mistakes developers are making in the 2026 job market is confusing a finished project with proof of engineering ability.
Y'all need to understand something:
Nobody is hiring you because your portfolio has a cool application. They ARE hiring you because of the engineering conversations that application allows you to have.
Think about two candidates walking into an interview with the exact same AI-powered project.
- Candidate #1 explains what the application does, which AI tools they used, and shows off the features.
- Candidate #2 explains WHY they designed it the way they did. Where they used deterministic logic instead of AI. How they handle unreliable model outputs. What happens when the system fails. What tradeoffs they made and WHY.
Same exact project. Two completely different interviews.
One demonstrated that they could produce software.
The other demonstrated that they understood how to engineer it.
And this is why I keep pushing developers to understand the decisions they're generating with AI.
I don't really care if AI wrote 100% of your code. But if you can't explain the architectural decisions, defend your tradeoffs, or identify where things could go wrong, you've missed one of the biggest opportunities that project was supposed to give you.
AI lowered the barrier to building software. It did NOT lower the bar for understanding it and making the right decisions.
We got into this on the latest episode of The Programming Podcast because I think developers need to fundamentally rethink what they're trying to accomplish with portfolio projects.
#code #coding #programming #softwareengineering
Danny Thompson retweeted
You should read and watch this...!!!
One of the biggest mistakes developers are making in the 2026 job market is confusing a finished project with proof of engineering ability.
Y'all need to understand something:
Nobody is hiring you because your portfolio has a cool application. They ARE hiring you because of the engineering conversations that application allows you to have.
Think about two candidates walking into an interview with the exact same AI-powered project.
- Candidate #1 explains what the application does, which AI tools they used, and shows off the features.
- Candidate #2 explains WHY they designed it the way they did. Where they used deterministic logic instead of AI. How they handle unreliable model outputs. What happens when the system fails. What tradeoffs they made and WHY.
Same exact project. Two completely different interviews.
One demonstrated that they could produce software.
The other demonstrated that they understood how to engineer it.
And this is why I keep pushing developers to understand the decisions they're generating with AI.
I don't really care if AI wrote 100% of your code. But if you can't explain the architectural decisions, defend your tradeoffs, or identify where things could go wrong, you've missed one of the biggest opportunities that project was supposed to give you.
AI lowered the barrier to building software. It did NOT lower the bar for understanding it and making the right decisions.
We got into this on the latest episode of The Programming Podcast because I think developers need to fundamentally rethink what they're trying to accomplish with portfolio projects.
#code #coding #programming #softwareengineering
Danny Thompson retweeted
💯 Judgements and decisions are now the moats
One of the biggest mistakes developers are making in the 2026 job market is confusing a finished project with proof of engineering ability.
Y'all need to understand something:
Nobody is hiring you because your portfolio has a cool application. They ARE hiring you because of the engineering conversations that application allows you to have.
Think about two candidates walking into an interview with the exact same AI-powered project.
- Candidate #1 explains what the application does, which AI tools they used, and shows off the features.
- Candidate #2 explains WHY they designed it the way they did. Where they used deterministic logic instead of AI. How they handle unreliable model outputs. What happens when the system fails. What tradeoffs they made and WHY.
Same exact project. Two completely different interviews.
One demonstrated that they could produce software.
The other demonstrated that they understood how to engineer it.
And this is why I keep pushing developers to understand the decisions they're generating with AI.
I don't really care if AI wrote 100% of your code. But if you can't explain the architectural decisions, defend your tradeoffs, or identify where things could go wrong, you've missed one of the biggest opportunities that project was supposed to give you.
AI lowered the barrier to building software. It did NOT lower the bar for understanding it and making the right decisions.
We got into this on the latest episode of The Programming Podcast because I think developers need to fundamentally rethink what they're trying to accomplish with portfolio projects.
#code #coding #programming #softwareengineering
One of the biggest mistakes developers are making in the 2026 job market is confusing a finished project with proof of engineering ability.
Y'all need to understand something:
Nobody is hiring you because your portfolio has a cool application. They ARE hiring you because of the engineering conversations that application allows you to have.
Think about two candidates walking into an interview with the exact same AI-powered project.
- Candidate #1 explains what the application does, which AI tools they used, and shows off the features.
- Candidate #2 explains WHY they designed it the way they did. Where they used deterministic logic instead of AI. How they handle unreliable model outputs. What happens when the system fails. What tradeoffs they made and WHY.
Same exact project. Two completely different interviews.
One demonstrated that they could produce software.
The other demonstrated that they understood how to engineer it.
And this is why I keep pushing developers to understand the decisions they're generating with AI.
I don't really care if AI wrote 100% of your code. But if you can't explain the architectural decisions, defend your tradeoffs, or identify where things could go wrong, you've missed one of the biggest opportunities that project was supposed to give you.
AI lowered the barrier to building software. It did NOT lower the bar for understanding it and making the right decisions.
We got into this on the latest episode of The Programming Podcast because I think developers need to fundamentally rethink what they're trying to accomplish with portfolio projects.
#code #coding #programming #softwareengineering
Danny Thompson retweeted
Learn how to become a better software engineer from an Apple veteran with over 12 years of experience building products.
This The Programming Podcast episode is with @RayFernando1337 and @leonnoel!
00:00 Meet Ray Fernando: 12 Years at Apple
01:51 How Ray Uses AI Agents to Build Software
05:24 The Real Cost of AI Agent Workflows
07:25 Automating Email with AI
10:52 A Word from Our Sponsors
13:13 Building Reliable AI Agents and Verification Systems
18:31 How to Keep Up with AI as a Software Engineer
21:01 Skill Chaining: Building Better AI Workflows
24:05 CSV vs. JSON and Why AI Constraints Matter
26:32 Claude vs. Codex vs. Gemini: Choosing the Right Model
31:17 Leon's AI Workflow for Shipping Code Faster
35:41 Behind the Scenes of Undersold AI
39:01 When Is an AI Product Ready to Launch?
43:04 Why Danny Walked Away from Two Software Projects
48:40 Fail, Pivot, Grow: Turning Failure into Progress
49:13 From Wedding DJ to Apple Engineer
53:09 How to Build Relationships and Network in Tech
58:19 What It Takes to Succeed as an Engineer at Apple
1:02:56 What Separates Great Engineers from Average Engineers?
1:06:50 Listener Q&A: What AI Portfolio Project Should You Build?
1:10:52 Why Most Developers Build the Wrong Portfolio Projects
1:14:41 Final Thoughts and Outro
Danny Thompson retweeted
Super hyped for this episode. Got into pstack & Grok Bot plus lots of alpha.
Learn how to become a better software engineer from an Apple veteran with over 12 years of experience building products.
This The Programming Podcast episode is with @RayFernando1337 and @leonnoel!
00:00 Meet Ray Fernando: 12 Years at Apple
01:51 How Ray Uses AI Agents to Build Software
05:24 The Real Cost of AI Agent Workflows
07:25 Automating Email with AI
10:52 A Word from Our Sponsors
13:13 Building Reliable AI Agents and Verification Systems
18:31 How to Keep Up with AI as a Software Engineer
21:01 Skill Chaining: Building Better AI Workflows
24:05 CSV vs. JSON and Why AI Constraints Matter
26:32 Claude vs. Codex vs. Gemini: Choosing the Right Model
31:17 Leon's AI Workflow for Shipping Code Faster
35:41 Behind the Scenes of Undersold AI
39:01 When Is an AI Product Ready to Launch?
43:04 Why Danny Walked Away from Two Software Projects
48:40 Fail, Pivot, Grow: Turning Failure into Progress
49:13 From Wedding DJ to Apple Engineer
53:09 How to Build Relationships and Network in Tech
58:19 What It Takes to Succeed as an Engineer at Apple
1:02:56 What Separates Great Engineers from Average Engineers?
1:06:50 Listener Q&A: What AI Portfolio Project Should You Build?
1:10:52 Why Most Developers Build the Wrong Portfolio Projects
1:14:41 Final Thoughts and Outro
Danny Thompson retweeted
I recommend watching The Programming Pod with @leonnoel @DThompsonDev and @RayFernando1337 talk about skill chaining, token efficiency and other cool agentic things 🔥🔥🔥 youtu.be/y-Kb3RSk0bo?si=SHsO…
Learn how to become a better software engineer from an Apple veteran with over 12 years of experience building products.
This The Programming Podcast episode is with @RayFernando1337 and @leonnoel!
00:00 Meet Ray Fernando: 12 Years at Apple
01:51 How Ray Uses AI Agents to Build Software
05:24 The Real Cost of AI Agent Workflows
07:25 Automating Email with AI
10:52 A Word from Our Sponsors
13:13 Building Reliable AI Agents and Verification Systems
18:31 How to Keep Up with AI as a Software Engineer
21:01 Skill Chaining: Building Better AI Workflows
24:05 CSV vs. JSON and Why AI Constraints Matter
26:32 Claude vs. Codex vs. Gemini: Choosing the Right Model
31:17 Leon's AI Workflow for Shipping Code Faster
35:41 Behind the Scenes of Undersold AI
39:01 When Is an AI Product Ready to Launch?
43:04 Why Danny Walked Away from Two Software Projects
48:40 Fail, Pivot, Grow: Turning Failure into Progress
49:13 From Wedding DJ to Apple Engineer
53:09 How to Build Relationships and Network in Tech
58:19 What It Takes to Succeed as an Engineer at Apple
1:02:56 What Separates Great Engineers from Average Engineers?
1:06:50 Listener Q&A: What AI Portfolio Project Should You Build?
1:10:52 Why Most Developers Build the Wrong Portfolio Projects
1:14:41 Final Thoughts and Outro
You can check out the podcast wherever you currently listen to podcasts!
Spotify
open.spotify.com/episode/4F6…
Youtube
youtube.com/watch?v=y-Kb3RSk…
Follow for more!
Danny Thompson retweeted
Will we still have a job in five years? On Nov 13, JetBrains AI.js Day puts frontend's scariest AI questions on air, live, online, and free.
Kent C. Dodds, Marina Wyss, Danny Thompson, and more.
Register now: lp.jetbrains.com/ai-js-day-2…
Danny Thompson retweeted
Be completely honest software developers. When is the last time you actually touched code instead of just prompting AI? – at Frisco, TX
43%in the last week
11%over a month ago
23%within the last 6 months
23%over a year
230 votes • Final resultsBe completely honest software developers. When is the last time you actually touched code instead of just prompting AI? – at Frisco, TX
43%in the last week
11%over a month ago
23%within the last 6 months
23%over a year
230 votes • Final resultsClosing keynote and a packed house after an extremely busy week at Team ’26 Europe in Amsterdam! From a Software Developer and Product perspective, this week was GOLD! The most valuable part of this week was proximity to problems.
Here are some huge takeaways:
- 15 MILLION+ MCP tool calls every day! Atlassian's MCP server just got a massive upgrade with 220 tools, expanding what developers and AI agents can access and accomplish across the Atlassian ecosystem.
- Loom can turn a screen recording into instructions for an AI agent. Instead of spending time explaining every visual change in a text prompt, developers can show what they want, talk through it, and have Loom generate structured instructions and acceptance criteria.
- Loom can also generate video walkthroughs of pull requests. It can run the previous and updated application, demonstrate what changed, explain implementation decisions, and show how the changes were tested. Think about how this changes code reviews as AI-generated PRs become more common.
- Atlassian built its new Artifacts application entirely on Forge! Not just a custom panel or extension. The UI, search, Teamwork Graph integration, Smart Links, and MCP tools were all built using Forge. If you've been wondering what you can actually build with Forge, this should expand your thinking.
- 250 BILLION+ objects and relationships in the Teamwork Graph. AI is only as useful as the context it can access. Connecting work, knowledge, code, people, and decisions gives agents a much better foundation to operate on.
- Up to 25% fewer tokens with the upgraded MCP server. More capabilities don't automatically have to mean higher token costs. How we design the infrastructure around our agents matters.
One thing became increasingly clear this week.
The conversation around AI development needs to move beyond which model you're using or how quickly you can generate code.
How are you giving agents the right context?
How are you making their work reviewable?
How are you connecting that work to the rest of your organization?
Those are engineering decisions, and they're becoming more important as AI adoption accelerates.
There's a LOT for Software Developers and builders to explore here!
#AtlassianTeam26Europe #Team26Europe #Atlassian
