Adrian 🧪 (@theadrianveidt) 's Twitter Profile
Adrian 🧪

@theadrianveidt

Full-time founder/entrepreneur
Part-time retail trader
Former Microsoft data scientist
Interested in memetics, psychology, health, & longevity

ID: 139596365

calendar_today03-05-2010 03:20:14

2,2K Tweet

875 Takipçi

1,1K Takip Edilen

Adrian 🧪 (@theadrianveidt) 's Twitter Profile Photo

Added a subscription option to my business this year (before we were one-time payment only) Recent subscriber growth has taken off, I think the key was aligning cadences more closely with user preferences eg by adding 3 & 6 month options

Added a subscription option to my business this year (before we were one-time payment only)

Recent subscriber growth has taken off, I think the key was aligning cadences more closely with user preferences eg by adding 3 & 6 month options
Adrian 🧪 (@theadrianveidt) 's Twitter Profile Photo

Update: canceled my Anthropic subscription in favor of Augment Code I wasn't using ClaudeCode enough to justify the expense In general I'm using AI tools less as I found they can create a tendency to be "intellectually lazy" vs thinking thru a clever design yourself

Adrian 🧪 (@theadrianveidt) 's Twitter Profile Photo

# of paid subscribers for comparison: - Netflix 300M - Amazon Prime 200M - Spotify 275M - Disney+ 125M - YouTube Premium & Music 125M - Apple Music 100M - Microsoft 365 Consumer 89M - Paramount+ 78M - Hulu 55M - Peacock 41M ⭐OpenAI 40M⭐ - ESPN+ 24M - NYT 10M - Peloton 3M+

Augment Code (@augmentcode) 's Twitter Profile Photo

Opus 4.5 is now live in Augment, and we're giving it to you at Sonnet prices for the next two weeks. We think it could become your daily driver. Try it today and let us know what you think!

Adrian 🧪 (@theadrianveidt) 's Twitter Profile Photo

My experience after buying Google's AI Pro Plan to try Gemini 3 for coding: I installed a vs code extension which made me install a cli which made me enable the preview features which sent me to a help web page which sent me to a google form which told me to go fuck myself

tae kim (@firstadopter) 's Twitter Profile Photo

Here's what's going to happen from here: -Nvidia will report blowout numbers the next several quarters on the NVL72 product supercycle (a step function up in capability with 72 GPUs in one AI server versus 8 GPUs). It will become one of the largest cycles in technology history,

Adrian 🧪 (@theadrianveidt) 's Twitter Profile Photo

Masterjohn’s conclusion is that for optimal longevity you should maintain elite level running performance. This corresponds to a 5k time of around 15-18 minutes. Past that point there’s not much value of trying to optimize vo2max or running performance.

Adrian 🧪 (@theadrianveidt) 's Twitter Profile Photo

AI coding tools are basically now the only software that we pay for at my company and we’re willing to pay a lot, bc they’re reaching a point where they’re good enough to replace all other software on demand.

Adrian 🧪 (@theadrianveidt) 's Twitter Profile Photo

Great Claude Code tips! I’ve been using subagents a lot more to great effect Subagents will burn a lot more tokens tho & have been responsible for me hitting my 5h 20x limit CLIs also unlock a lot of power eg can use GitHub’s to get agents to directly fetch more bug details