Instead of icebreakers, our school dropped us in the wild on day 1. Tomorrow we have to build a real AI tool for a local nonprofit By Friday, we need 1,000 followers and $5k in profit. We get 2 months to finish 20 side quests like this. Welcome to Day 1 @Alphaschool

Improving my developing skills with opus 5.5 The idea is a first person version of the viral mobile game: Stealth Master: Assassin Ninja

CYBERCAB WILL MAKE YOU LATE. If you need to get somewhere on time, do not pick a self-driving vehicle. Cybercab manages to choose the worst routes possible, turning 20 min rides into 40 min rides. I've taken Waymo and it struggles with the same issue, but Cybercab seems worse right now. They need to fix the routing ASAP. Outside of that, the actual car is unreal. The interior and ride comfort are on another level. A regular Uber or Waymo will never hit the same again. The screen and the sound system are both amazing. Price is another huge win. 95%+ of the time I've checked, Cybercab is cheaper than an Uber, usually by a substantial amount. The other glaring problem is availability. Demand is super high, especially on weekends when tourists come in. Sometimes pickup takes over an hour. You also cannot get a ride after 10pm, which keeps a lot of people from choosing it. I genuinely really like using it when I'm not in a rush, but the routing and wait times still have a long way to go.

Making a 1:1 copy of Clash Royale with @leo_tangg2. By day 4 I had AI agents running in a loop. They analyze hundreds of hours of real footage and play thousands of test matches to find glitches and improve the gameplay and visuals. Still a lot of work left, but it's getting closer.

Most founders waste months trying to build a unique app or find an undiscovered problem. There is a way larger chance of success if you find something that already makes money, take the core idea, and just market it better. Competitors shouldn't discourage you. They prove demand.

Running shoes that track distance, speed, and form built directly into the shoe. No watch to wear or charge, you just put them on and run. To anyone who runs: if the shoe itself was actually elite, would you ditch the watch?

Getting followers on X requires predictable value or entertainment. Which requires picking a lane. The 4 creator types: 1. The Contrarian: challenger energy with hot takes backed by something real 2. The Educator: frameworks and breakdowns of what actually works 3. The Operator: the CEO pov on what really happens behind the scenes 4. The Builder: building in public while sharing actual numbers, wins, and losses when you are starting out on X you should experiment to find out which lane you like posting about best, which lane you are best at posting about, and what lane will lead you to the most success.

Lebron James has the most misses in NBA history Over 15500 No one cares Most still consider him the GOAT It isn't about how many times you miss it only matters that you keep putting in the work, because eventually you will hit the shot.

It doesn’t have to be perfect for people to buy it. take a look at the spongebob ice cream.

Go ask ChatGPT to pick a number between 1-30. It will always pick 17. Every. single. time. Try it right now.

If you're paying for AI slop in 2026, you're a bum. I built my own protein tracker because I'm tired of paying for cal-ai slop. It took less than 2hrs and cost only $8.53. - Create > Consume - Build > Buy - Freedom > Control Now I'm subscription-free and have exactly what I want. At this point it's easier to make it than to pay for it. Just build.

benefits of lifting weights based on the amount of days you are going per week 1x/week - can compound to progress over time, way better than nothing 2x/week - big jump, enough to train major muscles 3x/week - good amount for most, can run full body 4x/week - ideal, lets you get peak out of volume and recovery 5x/week - good for more volume, but worse returns begin 6x/week - recovery becomes harder, usually more harm than good 7x/week - cooked, you are going to ruin yourself More days doesn't equal more gains. Recovery is when you actually grow.

If OpenAI plays their cards right, over the next few months they will become the clear favorite in the AI race. Though something we've all learned is that anything in this space can change in an instant