Man iphones have come a long way

My wife has worked in marketing for years and setting up photo shoots/video shoots has changed massively. A huge amount of the video and pictures that don't require a specialized lense are just done with feature phones these days.

You’re very good, I’m jealous of your work

Damn that’s clever

Many voices clone better than 11labs, while admitedly lower bitrate

Some of these don't feel 60fps, like the streaming one. I don't really know how to verify that though. Or maybe i'm just so used to 144fps.

Yes, otherwise there wouldn't be billions of dollars flowing into seed rounds every year to make novel products.

Not sure why you think a large company can copy a product in no time, they have to steer a cargo ship, you have a speed boat.

The fact that all metrics are relative doesn't suggest they got an amazing deal

Your notes aren't very good. They're not a time series database company, they're a columnar database company. But yeah the LLM bit is weird, database companies _always_ feel like charlatans when it comes to LLMs.

ClickHouse effectively has a number of personas. Time series is one of them, and ClickHouse has steadily absorbed market share from pure play time series databases over the last few years. Other personas include real-time observability backend (the single biggest use case in my experience) as well as real-time data lake engine. Time series support, column storage, and real-time response are key underlying capabilities. It's quite versatile and fun to use.

Disclosure: I run Altinity, a vendor in this space.

(Update: Disclaimer -> Disclosure. Sigh.)

Altinity isn't just a vendor, it's THE ClickHouse vendor before ClickHouse became a company. https://altinity.com/blog/big-news-in-the-clickhouse-communi...

always nice to see a database ceo be "one of us" and/or "write like a real human being".

That's very kind of you. We love working on ClickHouse and real-time analytics.

Willing to bet most columnar stores are used for time series.

I suspect most use of columnar databases is OLAP, which is different from what people usually mean when they say time series data.

I’d take that bet

saying they swear by the cursor composer model doesn't give me a ton of confidence

Checkpoints defined by the guest, not the host, was something I was working on recently too [1] (just before this came out). Very validating!

[1] https://github.com/danthegoodman1/checker