Peter Kao's Avatar

Peter Kao

@peterkao.com

• Founder at @MetanautLabs.com • 18+ years of tech entrepreneurship • 10+ years of running a VR studio • Current: https://risingtheory.com (fintech)

2,135
Followers
4,521
Following
512
Posts
06.10.2023
Joined
Posts Following

Latest posts by Peter Kao @peterkao.com

Ok so maybe AI can't replace a senior developer yet

27.02.2026 00:29 👍 0 🔁 1 💬 0 📌 0

Do you know how I know Meta juices their numbers? You can't unsubscribe to newsletters as a dev. The unsub link is there, the confirmation message is there, but they just never process it. This is a multi-billion dollar company and so I refuse to think this isn't intentional.

25.02.2026 18:09 👍 1 🔁 0 💬 0 📌 0

Alright, it took a couple of years but I know what my next project is going to be. Excited for the future 🚀

25.02.2026 00:24 👍 1 🔁 0 💬 0 📌 0

Anthropic is actually a better fit for Microsoft considering they lean more towards enterprise than OpenAI.

08.02.2026 02:50 👍 0 🔁 1 💬 0 📌 0

As a consequence of being in the commodities supercycle, countries around the world are going to see street crime tick up as copper, platinum, palladium skyrocket in value.

28.01.2026 17:40 👍 0 🔁 0 💬 0 📌 0

I should have gone to VRIC instead of VR conferences 😔

28.01.2026 04:48 👍 2 🔁 0 💬 0 📌 0
Preview
The Microstructure of Wealth Transfer in Prediction Markets Slot machines on the Las Vegas Strip return about 93 cents on the dollar. This is widely considered some of the worst odds in gambling. Yet on Kalshi, a CFTC-regulated prediction market, traders have…

Prediction markets can be gamed easily according to www.jbecker.dev/research/pre...

Just place limit orders and let others trade into you. Sell 'Yes' since people overpay for them. Focus on high-emotion, low-expertise categories like sports, entertainment, crypto.

21.01.2026 19:20 👍 1 🔁 0 💬 0 📌 0

Wealthsimple is the closest thing to a super app in Canada.

20.01.2026 00:03 👍 0 🔁 0 💬 0 📌 0

AI diagnosis tools are going to be crucial in debugging AI-controlled robots. Robots process billions of parameters that human technicians aren't able to diagnose. Someone work on this, stat.

15.01.2026 23:07 👍 0 🔁 0 💬 0 📌 0

Video clips are today's pamphlets.

15.01.2026 21:56 👍 0 🔁 0 💬 0 📌 0

Content creators and online personalities have started sharing their portfolios' performances for 2025. Seeing them barely beat the S&P500 (and with so much effort) while being so confident in their posts strengthens my conviction that you have to do the research yourself.

11.01.2026 19:46 👍 0 🔁 0 💬 0 📌 0

I've always wondered: how well do venture capitalists perform on the stock market?

27.11.2025 06:00 👍 0 🔁 0 💬 0 📌 0

When people ask entrepreneurs how many employees they have, what they're actually asking is how successful are you. This question will be outdated in a few years as smaller companies with higher revenues and lower profit margins (compared to non-AI software businesses) become the norm.

10.11.2025 18:21 👍 1 🔁 0 💬 0 📌 0

The 21st century play is turning out to be selling the materials needed to sell the shovels.

13.10.2025 01:20 👍 2 🔁 0 💬 0 📌 0

Could retail stock participation decline if tech weakens while precious metals and rare earths outperform? 🤔

09.10.2025 22:54 👍 0 🔁 0 💬 0 📌 0

Canada's economy is going to be saved by mining. It's a great time to look into shifting sectors in your portfolio if you haven't done so already.

07.10.2025 19:50 👍 0 🔁 0 💬 0 📌 0

Over a decade ago, I traveled the world as a digital nomad running an ecommerce business. If I were to do it again today, trading and investing in the stock market would probably earn me more money than ecommerce. Crazy times.

06.10.2025 19:55 👍 3 🔁 0 💬 0 📌 0

I've been "fact checked" by people using AI so many times now. It's crazy to me how people still don't understand how LLMs work. And it's even crazier to think how this will develop poor and intellectually lazy habits in people, en masse.

05.10.2025 16:21 👍 2 🔁 0 💬 0 📌 0

Gemini models still sound like a robot. If Google is after the consumer market, they need to make it sound more like OpenAI, or better, Anthropic's models.

03.10.2025 16:54 👍 0 🔁 0 💬 0 📌 0
Post image

AI will be the greatest bundler of our century.

02.10.2025 20:23 👍 2 🔁 0 💬 0 📌 0
Post image

No one's going to buy your product if you use poorly AI generated images like this. AI generation is fine but don't be lazy about it.

22.09.2025 20:08 👍 1 🔁 0 💬 0 📌 0

Wow Codex is actually better at programming than Claude Code right now. How fast things change!

19.09.2025 05:34 👍 1 🔁 0 💬 0 📌 0
Post image

Due to the "lost in the middle" issue with current LLMs, you should put the most important parts of a long message to LLMs at the beginning as per the research paper, "Lost in the Middle: How Language Models Use Long Contexts" (arxiv.org/abs/2307.03172 )

18.09.2025 19:49 👍 3 🔁 0 💬 0 📌 1

I get why @sama (Sam Altman) don't capitalize words anymore. LLMs don't care for capitalizations. When you use LLMs for so long, it starts to affect how you write.

16.09.2025 20:18 👍 0 🔁 0 💬 0 📌 0
Post image

Made this with my daughter over the weekend. Her face lit up when she realized she could bring her ideas to life with AI just by describing what she wanted to see. What a great way to teach kids how to be more detail-oriented.

15.09.2025 20:02 👍 1 🔁 0 💬 0 📌 0
Post image

Has the world gone insane? Headlines in my inbox today.

15.09.2025 17:20 👍 0 🔁 0 💬 0 📌 0

Favorite thing to do with my daughter right now: Creating AI generated images in the style of Cocomelon. It's a great way to teach her to think and describe things in detail while preparing her for the inevitable AI future.

15.09.2025 00:16 👍 0 🔁 0 💬 0 📌 0
Video thumbnail

One of my favorite standups. Happy Friday everyone

05.09.2025 15:46 👍 1 🔁 1 💬 0 📌 0

It's crazy how much Sonnet and Opus have declined in quality, and it only took one month. This is another friendly reminder to never pay for annual subscriptions when it comes to AI models.

05.09.2025 00:08 👍 0 🔁 0 💬 0 📌 0

Seeing how people write prompts on X (grok) is alarming. Some of these prompts are terribly framed with no context. If this is how the average person writes prompts, future chat interfaces for all AI should show a list of questions to ask next like predictive autocomplete on kb

04.09.2025 16:31 👍 0 🔁 0 💬 0 📌 0