datadrivenangel

3 ms·

1.
by datadrivenangel 2mo ago
and the fun question of how do you find good problems is very hard to systematize!
2.
by datadrivenangel 2mo ago
Why would they use Notion in a system like this? Seems like setting yourself up to have a massive pile of janky cruft.
3.
by datadrivenangel 2mo ago
Congrats on the relaunch! why should I use this versus luma or something similar?
4.
by datadrivenangel 2mo ago
One of the lawyers here got put on probation a decade ago for dumping a client and saying: ""[P]laintiff’s opposition is absolutely untrue except that I am prejudiced against her. I despise her and would rather shovel manure than
5.
by datadrivenangel 2mo ago
Actual legal document: https://drive.google.com/file/d/1ufb8Bg9RA9LmRKOvXSm9UbP9YIT...
6.
by datadrivenangel 2mo ago
like this one? https://eschrock.dtrace.org/2008/07/
7.
by datadrivenangel 2mo ago
I cancelled my subscription over this. [0] 0 - https://www.williamangel.net/blog/2026/09/18/i-cancelled-my-...
8.
by datadrivenangel 2mo ago
This allows us to remove our one line Claude.md files that just say "AGENTS.MD"
9.
by datadrivenangel 2mo ago
I actually like this as a method for getting over writers block. The second the AI starts spitting out words I know it will be wrong and dumb and so I might as well do it myself. Enough activation energy to get started
10.
Claude Code now reads AGENTS.md if there is no Claude.md(code.claude.com)
734 points by datadrivenangel 2mo ago|275 comments
11.
I Cancelled My Claude Subscription(williamangel.net)
6 points by datadrivenangel 2mo ago|3 comments
12.
by datadrivenangel 2mo ago
It's funny because it's so true
13.
by datadrivenangel 2mo ago
Different flavors of magic require different ritual components and in the name of all that is stable and production worthy, please don't get a necromancer to do your civil construction magic. Vendor-neutrality helps reduce lock-in, and
14.
by datadrivenangel 2mo ago
Absolutely not. This is inherently insecure.
15.
by datadrivenangel 2mo ago
And bash is fast! Reminds me of the 230x faster than hadoop blog: https://adamdrake.com/command-line-tools-can-be-235x-faster-...
16.
by datadrivenangel 2mo ago
Shareholder nuisance lawsuits can happen basically any time the stock goes down... or doesn't go up enough, and often companies will settle instead of the expensive fighting...
17.
by datadrivenangel 2mo ago
If the code is simple, the tests aren't much extra work. High test coverage doesn't mean your code is good, but it at least reduces the rate at which you accidentally break stuff.
18.
by datadrivenangel 2mo ago
SLAs don't relly get you meaningful money back, they give you a way to get out of long contracts early and switch to a better vendor.
19.
by datadrivenangel 2mo ago
Dirty reads is a classic symptom of scaling poorly. Eventually consistent database writes are fine, but it's a bad look if a user takes an action and then doesn't see the results of their action.
20.
by datadrivenangel 2mo ago
but it's bad enough to be worth not upgrading.
21.
by datadrivenangel 2mo ago
non GAAP. If they were GAAP profitable they either wouldn't be sharing or would be IPOing already.
22.
by datadrivenangel 2mo ago
With better ergonomics around handling failures.
23.
by datadrivenangel 2mo ago
Open weight models are all built on IP theft as well. Also more like 6-12 months behind, maybe 18 but it's hard to say.
24.
by datadrivenangel 2mo ago
LiteLLM's problem isn't really features, it's how bloated all the features are, and specifically how AI maximalist and janky their dev practices are.
25.
by datadrivenangel 2mo ago
if it was less elegant from a function/cost perspective, it would have been replaced a long time ago.
26.
by datadrivenangel 2mo ago
Gattica disagrees.
27.
by datadrivenangel 2mo ago
humanoid robots are decades away, and they are not going to be good or cheap for a long time.
28.
by datadrivenangel 2mo ago
so for an hour flight it costs $5?
29.
by datadrivenangel 2mo ago
Data Science Tasks (Internal) doesn't include time for Astra... same for Database Migration Tasks (Internal)... But does for gpt 5.6 sol.... which is funny. Same for HealthBench Professional and a few others. Clearly either OpenAI is v
30.
by datadrivenangel 2mo ago
150k tokens per minute at 1.5k tokens per second means you can have like 3 users concurrently and that's not a lot.

More ›