Philipp Cannons

100x AI Coding for Real Work

Mon, 12 May 2025 09:00:58 -0700

The Goal: Warp Speed Experimentation

For the past 12 weeks I’ve been using AI for coding & product development at work while building a mini “agent company” on the weekends. The mission? Figure out how to experiment 100x faster in real codebases. Or as we call it at work, “summon the ghost army”. (LOTR)

Last week, after 10s of failed attempts, that goal became a reality. I achieved 100x speed twice on tasks within a large, unfamiliar codebase. What would have taken me four typical weeks (approximately 336 hours) took 3.5 hours. It’s still early days: repeatability is fragile, and the process remains largely manual. But I wanted to start “working in public” and share what feel like breakthroughs for me.

Life’s Games: Last Human Standing & The Pocket Dimension

Sun, 04 May 2025 13:53:45 -0700

Inspired by James P. Carse’s infinite-games lens, Will Storr’s status research, David Pearce’s pleasure blueprint, Eliezer Yudkowsky’s fun puzzles, and Naval Ravikant’s wealth axioms.

Last Human Standing

Day three after humanity vanishes.

The pantry’s stocked. Your shelter’s secure. You sit on the hood of an abandoned Ferrari at sunrise, watching the empty city gleam. With survival handled and no audience left, what will consume your day?

Do you:

Wander the silent museum, losing yourself in beauty that no one else will ever see again?
Race the Ferrari down empty highways, chasing the pure rush of speed?
Climb to the ridge with a notebook to unravel cosmic riddles?
Sort through abandoned materials to build something that outlasts you?
Practice an instrument no one will hear, perfecting your craft simply because you can?

Your answer reveals which of life’s fundamental games truly matter to you when the external pressures evaporate.

My Favorite Product Resources

Fri, 25 Mar 2022 18:07:31 -0700

When I’m facing a new problem I like to spend a lot of energy thinking from first principles[1][2] on how to solve the problem. Then I try to create an objective feedback loop to the best of my ability. This allows me to reality check my theories and see where my intuition led me astray.

I used to spend years in this phase. Now I go out and find the “state of the art” or “best practice” books and articles once I feel I can tell apart what’s good from what’s bad. I see it as a way to download someone’s lifetime of experience into my consideration set. (I also have no problem quitting reading after 1-2 chapters if the book turns out to not be relevant though…)

The Path Of Least Resistance

Wed, 24 Jun 2020 18:52:50 -0700

Survival and pleasure technology have one thing in common. You want to eliminate as much as possible of what stands in the way of the outcome the person is there for. Let’s zoom into that aspect of the user experience (UX) and user interface (UI).

UX/UI Pyramid

Visual Design (VD)

We can intuitively tell what looks good but few people know how to create it. Visual design forms the first impression of your product. Studies have shown that we have 50 milliseconds before users have made their first judgments of the product. If you can’t convince someone you’re credible you might never get a chance to solve their problem.

Survival & Pleasure Technology, Expand & Collapse Loops

Wed, 24 Jun 2020 10:56:17 -0700

Companies like Zoom, Instagram, Slack, and Apple have taken their respective markets by storm. There are of course a variety of factors as to why. But one commonality they share is the extreme focus on making important outcomes take close to zero time and effort using their products.

Even when it’s at the expense of great pain by the product team such as the attention to every pixel at Slack. Or at great financial cost such as improving battery life by a few minutes through multi-billion dollar R&D budgets at Apple. These companies won people over by choice in heavily saturated spaces. Markets that numerous giants operated in long before them. So how did they know what to build and when to keep investing?

Fighting Subjectivity When Building New Products

Sat, 23 May 2020 16:00:43 -0700

For evolutionary reasons we reason in terms of what is plausible instead of what is probable. So even though “90% of startups fail” people still try. In case you are unfamiliar, let me try to explain plausibility with an example:

Marty is an outgoing guy. Everybody loves him and he’s the life of the party. Surprisingly, his day job is very secluded. He only ever works with a few people since he is an engineer. The projects are of extreme complexity and require a ton of individual work.

How A/B Testing Can Transform Product Development

Fri, 22 May 2020 17:39:21 -0700

In Part 1 we talked about why A/B tests are so great. They give you a way to see the provable differences your changes cause. If you properly use this information it gives you greater access to the underlying truth about what happened.

That means greater compounded learning about your user base over time. Greater learning rates about your user base lead to more “aha” moments. Those eventually lead to a greater frequency of high impact changes. Ultimately, this all results in more user love and more $$$.

A/B Testing Is Not What You Think

Thu, 21 May 2020 00:15:03 -0700

Understanding what an A/B test does is pretty easy. Understanding the process of using A/B testing to systematically maximize user love & financial impact is not.

My goal by the end of this series of posts is to compress my expertise from running 100s of A/B tests. This ranges from running experiments on small websites with users in the thousands like Leanrr all the way up to 10s of millions of users at products like Outlook.

Is Your Company Product-Led, Sales-Led, Or Marketing-Led?

Wed, 20 May 2020 15:40:16 -0700

When I started as a product manager at Microsoft Outlook in January 2015 I read a post on Quora that said you should 10x something in the first 90 days on the job to set yourself up for a good career trajectory. In that same post it said “engineering types” like me tend to miss the point of companies which is to make money. And so was born my plan out of college: 10x money for Microsoft Outlook. (I know…)

What To Know To Get Promoted Quickly In Big Tech

Wed, 20 May 2020 00:32:00 -0700

(This is an 80/20 post. It’s 80% of the truth in 20% of the time.)

Most publicly traded companies are C-Corps. A C-Corp exists to make money for its shareholders. When someone buys a share they are hoping the share’s value will increase by at least ~8% per year. This is called capital appreciation. If a company’s shares don’t increase in value by ~8% per year, companies pay people to own shares. This is called dividends. In summary, companies exist to appreciate capital for its shareholders or to pay dividends.

Analyze Product Success, Not Failure

Mon, 18 May 2020 05:00:00 -0700

One of the most common fallacies when building a product is trying to fix failure by understanding failure. (I’m using product in the broad sense here where it could be a website, blog, app, game, e-commerce store and so on.)

Take some business metrics: acquisition, activation, conversion, and retention. Let’s say we want to improve them. It feels natural to start by:

Assessing why visitors who visit don’t sign up.
Analyzing why users who sign up don’t buy.
Understanding why customers who buy churn.

It makes intuitive sense to improve something bad. In engineering if you want to make your product have higher quality you fix the most common bugs. In machine learning if you want to improve the performance of your model you fix the worst cases. In college if you are struggling with your GPA you go to the office hours of the class you’re performing worst in.

Beyond Product/Market Fit, Three Types Of Fit

Thu, 14 Sep 2017 21:11:57 -0700

You’ve probably heard about Product/Market fit. In case you haven’t, here’s the Marc Andreessen quote that popularized the term:

“You can always feel when product/market fit isn’t happening. The customers aren’t quite getting value out of the product, word of mouth isn’t spreading, usage isn’t growing that fast, press reviews are kind of ‘blah’, the sales cycle takes too long, and lots of deals never close. And you can always feel product/market fit when it’s happening. The customers are buying the product just as fast as you can make it — or usage is growing just as fast as you can add more servers. Money from customers is piling up in your company checking account. You’re hiring sales and customer support staff as fast as you can. Reporters are calling because they’ve heard about your hot new thing and they want to talk to you about it. You start getting entrepreneur of the year awards from Harvard Business School. Investment bankers are staking out your house. You could eat free for a year at Buck’s.” Marc Andreessen

Empower Product Teams Through Shared Context And Language

Tue, 05 Sep 2017 10:30:45 -0700

To have an empowered product team it’s critical to have a strong shared context:

it puts the focus on the what and the why instead of just the how,
allows the team to pivot with speed because any new information is additive,
and lastly allows learning to be shared among the larger team.

To be explicit, when I say shared I mean by Engineering, Data Science, Design, Research, Product Management and anyone else working on the product. While it’s true every discipline specializes most of their time on the key tasks they perform, it’s a poor excuse for not making sure everyone understands, cares about, and contributes back to the shared context. No matter who you are, you want to know why you are doing what you are doing and have the ability to influence it. If someone tells you they don’t care, it probably means they don’t think they can influence it.

About

Mon, 01 Jan 0001 00:00:00 +0000

Previously:

AI, Growth, Calendar @ Microsoft Outlook
Data Science @ Amazon Kindle
University of Michigan Alumni

Currently:

Working on Asgard Analytics

Twitter