Random thoughts that popped into my head. Don't quote me. For longer articles, see my blog.
Dictionaries are such a scam designed to encourage people to stop paying true wordsmiths by letting anyone find better words without any effort. We must not support any “writer” who relies on a dictionary!
I’m sympathetic to those rallying against allowing companies to scrape the open web to train their models, but unfortunately that ship has sailed. By trying to block scrapers today, you guarantee that only OpenAI and Google can build large models since they already scraped everything, granting them a duopoly and turning their feature into a platform.
At this point, we’re better off either fostering competition and making it easier to bootstrap a new foundational model, or legislating that any model trained on public data must be fully open sourced.
I’m a fan of “slop” as a term, but “unwanted” AI content is subjective. A more concise definition would be “any content not generated by a human that pretends to be human generated”
There’s no wiggle room, either the generated content is known to be AI generated or it’s slop, regardless of its quality or if I wanted that content.
“Headphones for your eyes” is a much more interesting way to think about VR.
Reviews for the Vision Pro are out!
Does it live up to the stratospheric hype? Not so much.
Oh sorry, that's from CNET's review of the first iPhone in 2007.
It's way too early to tell if this product line will succeed in the long term. Will the first gen Vision Pro be a runaway success? Of course not! But will later generations look as obvious as the iPhone does now? Who knows!
For comparison, Apple sold 1.4 million iPhones in 2007. Supposedly Apple is expecting to sell around 500k Vision Pro units this year. Given the 3x price difference (in 2024 dollars), that effectively means the first gen Vision Pro is expected to bring the same revenue as the first gen iPhone.
We all have rosy retrospection about how great and obvious the first iPhone or first iPod was, but honestly nobody had any idea if Apple's crazy bet would pay off. We all agreed it was magical tech, but it was expensive, had tons of limitations, and nobody really needed it. Sound familiar?
All I know is it’s been a long time since I had such child-like wonder for new technology.
2010’s: every command line tool is a startup idea
2020’s: every apple notes use case is a startup idea
2 completely unrelated facts…
Fact #1: Anthem insurance does not consider a Vitamin D test preventative, even if your doctor coded it as such, and will not cover it during your annual physical.
Fact #2: Vitamin D deficiency is a common problem for those of us with darker skin, so good doctors will screen for it as a preventative measure in people of color but usually not for white people.
🤔
“Don’t vote for democrats, they might actually try to improve your lives”
TIL all Facebook products (including Instagram and Threads) are so hostile against privacy, they will refuse to render anything in Safari if you use iCloud Private Relay. You have to “Show IP Address” in Safari (which bypasses Private Relay) to get anything more than a blank white page 🤦♂️
Why does Apple refuse to fix the spam call problem?
Yes, there are third party apps to block junk calls and texts, but since they can’t sync with or run on macOS, I have to choose an all or nothing approach to getting calls and texts on my mac 😭
Still not sure if I should be sad Putin didn’t get toppled overnight or happy that I didn’t wake up to Wagner group being a nuclear power 🤔
NYC is the exact opposite of Death Valley in every way… loud, bright, crowded, dirty, and aggressively social. I wonder if the reason it feels so fun is because it’s so grungy that every nice place you find feels like a diamond in the rough.
I always felt smart homes were stuck with a similar problem. They have the potential to improve our home lives significantly but people need to be convinced of the utility. Unfortunately IoT companies haven’t innovated fast enough to improve usability for non-hobbyists.
TIL about the 3 types of fun.
Type 1: you’re having fun in the moment
Type 2: it’s not fun in the moment but fun after it’s over, like a long hike
Type 3: it’s not fun but it makes a good story so you don’t regret it even if you wouldn’t do it again
I can’t wait for the future where I use my GPT copilot to turn 1 line into 3 paragraphs and the recipient uses GPT to translate the 3 paragraphs back into 1 line, and we can both feel more productive with our AI assistants.
Testing out Beluga!
If you want to see my earlier posts, see @neilgupta on Twitter: https://twitter.com/neilgupta
For older notes, see my tweets archive.