Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
qudent
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
3 ms
·
1.
▲
Anthropic, Cypherpunks, and the Bomb: 3 Rounds of Technologists vs. the State
(github.com)
1 points
by
qudent
7mo ago
|
0 comments
2.
▲
by
qudent
8mo ago
I think the fact that it is not possible to put hard spending caps on API keys might be ruled illegal by some EU court soon enough, at least when they sell to consumers (given the explosion of vibecoding end-users making some apps). When I
3.
▲
by
qudent
8mo ago
In Google AI Studio, Google documentation encourages to deploy vibecoded apps with an open proxy that allow equivalent AI billing abuse - giving the impression that the API key were secure because it is behind a proxy. Even an app with 0 AI
4.
▲
Tell HN: Google AI Studio docs encourage Google-discoverable open wallets
(github.com)
1 points
by
qudent
8mo ago
|
1 comments
5.
▲
by
qudent
8mo ago
Google AI Studio documentation encourages developers to deploy vibecoded apps, claiming the API key is secure because it is protected by a proxy - however, there are no checks on the open proxy the deployed app exposes, which allows anyone
6.
▲
by
qudent
8mo ago
Because inference is autoregressive (token n is an input for predicting token n+1), the forward pass for token n+1 cannot start until token n is complete. For a single stream, throughput is the inverse of latency (T = 1/L). Consequentl
7.
▲
by
qudent
8mo ago
It does affect the throughput for an individual user because you need all output tokens up to n to generate output token n+1
8.
▲
Google AI Studio's API key protection is as exposed as the key itself
(github.com)
2 points
by
qudent
9mo ago
|
2 comments
9.
▲
by
qudent
9mo ago
Google AI Studio's Build Mode hides API keys behind a proxy during deployment, which the docs imply is secure. But the proxy forwards arbitrary requests to any Gemini model with zero auth, quota or validation, using the real API schema
10.
▲
by
qudent
9mo ago
The "correct" way to have agents working on repositories in parallel - which you want to do to speed up AI development rather than just waiting around - is to use git worktrees, which are separate directories checking out differen
11.
▲
Show HN: Claude skill+design pattern for managing worktrees for parallel agents
(github.com)
2 points
by
qudent
9mo ago
|
1 comments