When I first saw Claude generating PPT decks by writing Python code instead of making the XML directly, it was sort of an "aha moment" for me. This seems to be the path for many things. It also feels slightly limiting, and like a hack LONG term, but 100% correct approach for a while.
This isn't sustainable.
Arguably margins were too low is some parts of semiconductor space, but we still need lower cost per transistor over time or were in trouble.
I've taken to buying SACDs when possible. The format supports higher dynamic range, but that barely matters. The mix is the bigger issue and SACD mixes are often better.
Note you need an SACD player.
And also note this only applies for playing on a proper HiFi or with good headphones at least.
In your car, etc you probably want the compressed mix.
I doubt an LLM is calculating withholding. I presume 99.9% of the actual logic will still execute in QuickBooks or Paychex etc. Lots of this sounds like cross system orchestration against well defined APIs.
Yes, there's still danger, it could use the APIs wrong, but humans can use the GUI wrong too
I work at a less innovative place, and I see out product managers coming with prototypes, at least solid mock ups rather than just a jira.
They socialize it with potential users, they iterate, they find missing requirements, it's pretty powerful.
The net result is we're building better features faster.
My comment history here is full of complaints about MS Teams, the chat app.
It suffers from the "re-use every existing MS tech" problem. Building it on top of SharePoint I'm pretty sure resulted in it's top problems over the years (some fixed):
- search sucked
- can't scroll back to old messages
- couldn't do private channels
- the very concept of 1:1 teams to SharePoint site, resulting in a millions teams when all you really wanted was a channel
- can't rename teams or channels
- couldn't do private channels
I'm sure many more I didn't catch.
These are all observations from outside, I've never worked at MS