Anthony Nowell (@atnowell) header
Anthony Nowell (@atnowell) profile picture

Build stuff. Care for people. https://teapawt.net/t.co/erKa3ttdEr

Seattle
Joined March 2009
Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)Photo by Anthony Nowell (@atnowell)
Trust is slowly earned and quickly lost — applies in software, "tech", and especially AI.
I spend a lot of time around people who have never worked in software. Reading Twitter, you'd never know how much those people currently hate software, "tech", and especially AI. Maybe the Twitter bubble doesn't care. But if we're not building for them, who are we building for?
1
100
Interviewer: Implement FizzBuzz. Me: Ok. Let's spin up Claude, grab a Jev API key, and do some evals...
1
1
56
Don't be a meat proxy. Best read of the week.
54
Anthony Nowell retweeted
If your AI writes the word "honest", immediately kill the session and throw away whatever it produced.
22
12
2
133
18,400
Anthony Nowell retweeted
The Certifiably Insane Way to Build an AI Agent: 1. choose a category where mistake tolerance is roughly the same as it is in self-driving cars. we chose "email-based scheduling assistant." many people want this product, but they immediately fire him if he screws up an interaction with a prospect, a candidate, or a potential investor 2. you learn that the edge cases are too complex and too frequent to be solvable. ours: managing timezones for people who travel (and change travel plans) constantly. knowing when NOT to respond, when to text the customer on the side to verify something, when to follow up, which sub-calendar to use, when to bend the rules on availability, when we can schedule that one type of call during your commute but not the other type of call. sharing your availabilities without compromising your privacy. and on and on. 3. the product doesn't feel viable, but you don't want to give up. you spend hours in a hot tub in Marin with a friend who makes self-driving cars. you make a plan to do it the way they did: hold the steering wheel. you go home and build a human-in-the-loop platform and hire contractors to serve as a backstop and catch mistakes before they happen (and to help design a map of what a world-class EA would do in every weird scenario). you decide trust is the currency in your category, so it must be the thing you won't compromise on. the product must succeed at any scheduling request, no matter how complicated. 4. you instantly feel an overwhelming market pull. so you keep going, growing that team to 75 people working 24/7 to support the nonstop scheduling needs of your customers. tons of engineering time goes to scaling the human platform instead of building the product. 5. you try to raise a Series A and investors say you are insane. your gross margins are extremely negative. they believe this is a problem worth solving, but they don't believe it is as hard to solve as you say. they want AI, not humans. your competitors put "NO HUMANS IN THE LOOP" on their landing pages to call you out. you keep going. 6. you work day and night building the harness that can meet the quality standard your customers have come to expect. you create a massive synthetic gold dataset. audit it, and clean it, label it. repeat. then, experiments. fine-tuning. RL. ACE. DSPy. sub-agents. sub agents for your sub-agents. rebuild the harness. throw more tokens at the problem. 7. some weeks you make big progress. some weeks your evals climb a single basis point, but that's better than nothing. more experiments. more tokens. john coogan said the hot trend in 2026 will be dogged pursuits. that pushes you to continue the pursuit, doggedly. 8. then, one day, you realize you are scheduling thousands of meetings a day and approaching 50% autopilot with no increase in churn or complaints. you put 150 customers in a full self-driving experiment, and they use the product MORE than they were using it when they had the human backstop. you can really start to let go of the steering wheel. 9. you don't know yet if this was a hill worth climbing, but you are nonetheless stoked that you can see the top. you have created a proprietary map of what to do in a million different situations. nobody else has that map, and the models keep getting better at following maps. your plan was to bet on trust, and your product can be trusted. today was the first day Howie crossed 50% autopilot:
77
67
30
942
320,138
Anthony Nowell retweeted
No comments.
172
3,143
187
25,246
973,914
Artificial Incompetence - I'm undecided if it's better or worse than Human Incompetence.
29
no lunch break until Claude Max limit is reached: ✅
57
Anthony Nowell retweeted
I've made ads for the Super Bowl before. Spent millions to make them and air them during the Big Game. But this one -- this is the one that made the most impact. Zero media dollars. 40m views. Because it speaks the truth. Ban us all you want. Truth outs in the end.
900
15,200
541
75,529
2,828,234
Move over IoT - the S in AI is for security
22
Anthony Nowell retweeted
Meet Howie, the people's secretary. Out now. (see thread for signup link + fundraise announcement)
101
58
70
1,205
1,490,174
Anthony Nowell retweeted
It’s all about Jesus.
7,332
108,240
5,096
406,769
25,645,690
Reported my first-ever security bug to @SlackHQ. Disappointingly closed: "works as intended." Fwiw, think twice before adding collaborators on a Slack app intended for supporting local development. They may be able to read your DMs. Since it's closed: gist.github.com/anowell/5176…
85
My team has been growing. Ralph's productivity is underwhelming, but he's still a critical part of the team.
1
1
58
Gus, my Golang expert was fired, but I might be adding a Flutter dev soon.
42
Dear @OpenAi - maybe fix your ChatGPT memory leak. Pretty please?
1
161
Anthony Nowell retweeted
You can fix your existential crisis in under a month. Here's how: All too often, people who worked for me would begin to drift. - Energy wanes. - Performance drops. - Motivation disappears. Then I'd hear the dreaded words: "I just don't know what I'm doing with my life." As we start to unpack that crisis, I usually discover a twisted knot of emotions, dreams, and fears. Real and imagined. Personal and professional. Instead of small, solvable problems, they rolled them up into one giant tangle. The existential crisis. After much trial and error, I found a simple but foolproof way to help them find their path. ----- 1️⃣ Split a piece of paper into three columns: - Questions - Action - Answers For the Answers section, use the four "Ikigai" subheadings - I love... - I'm good at... - I can get paid for... - The world needs... 2️⃣ Dump all the questions and answers out of your brain Answers - These are the things you already know. There is value in reminding yourself what those are. It's easy to lose sight of them in the storm. Questions - These are the things you're wondering. Or anything is causing you angst. Just empty your mind. No question is too small. 3️⃣ Prioritize the questions Which one do you think is most valuable to answer? Don't overthink it. Follow your gut. Move this to the action column and write down the simplest thing you can do to answer that question. Which one is causing you the most angst? Write that one down, too. Again, identify the simplest action that will give you the data you need. If you're stuck, the action section can be a great place to get help from a mentor, coach, advisor, or partner. 4️⃣ Take action This is the part that matters. You have to take the action you commit to every day. Have the conversation. Write the journal entry. Whatever it is, do it. Once you take that action, you either answer the question and write it in the Answer column. Or you learned what the next action needs to be in order to answer it. Write that down and commit to doing it tomorrow. 5️⃣ Repeat Once you've answered a question, go back to your question column and pick the next one to answer. ----- That's it. I ask my team to commit to doing this for one month. No one has ever needed the full 30 days to gain the clarity to move forward. It works...if you do the work.  - You have to write it all down.  - You have to take the committed action. I've used this so often that I built a simple template.  Subscribe to the MGMT Playbook for free access. mgmt.beehiiv.com/p/simple-bu… You get this template and dozens of others free. Before you go: -> Please repost ♻️ it to your network. -> And follow @dklineii for more.
9
17
154
44,257