Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
dbreunig
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
31.
▲
by
dbreunig
7mo ago
Last year they pushed out an update stating if any “Meta AI” is left on, they can access image data for training, I turned the AI off and used them as headphones and taking videos while biking. After a couple rides, I couldn’t bring myself
32.
▲
Claude and the Dow: AI is unlike other tech because AI has embedded judgment
(dbreunig.com)
1 points
by
dbreunig
7mo ago
|
1 comments
33.
▲
by
dbreunig
7mo ago
Check out “Recursive Language Models”, or RLMs. I believe this method works well because it turns a long context problem (hard for LLMs) into a coding and reasoning problem (much better!). You’re leveraging the last 18 months of coding RL b
34.
▲
Two Beliefs About Coding Agents: Devs Don't Realize What They Bring
(dbreunig.com)
2 points
by
dbreunig
7mo ago
|
0 comments
35.
▲
by
dbreunig
8mo ago
Author of the post here. I didn’t say AI was bad and I acknowledged the benefits of Electron and why it makes sense to choose it. With 64gb of RAM on my Mac Studio, Claude desktop is still slow! Good Electron apps exist, it’s just an intere
36.
▲
by
dbreunig
8mo ago
I keep saying this, it’s my new favorite metaphor.
37.
▲
Why is Claude an Electron app?
(dbreunig.com)
428 points
by
dbreunig
8mo ago
|
458 comments
38.
▲
Analyzing How System Prompts Define Agent Behavior
(dbreunig.com)
3 points
by
dbreunig
8mo ago
|
0 comments
39.
▲
The Potential of RLMs
(dbreunig.com)
3 points
by
dbreunig
8mo ago
|
1 comments
40.
▲
The Rise (and Limits) of Spec Driven Development
(dbreunig.com)
3 points
by
dbreunig
8mo ago
|
0 comments
41.
▲
by
dbreunig
9mo ago
That's cute.
42.
▲
A OSS Library with No Code, Only Specs
(dbreunig.com)
3 points
by
dbreunig
9mo ago
|
2 comments
43.
▲
2025 in Review: Jagged Intelligence Becomes a Fault Line
(dbreunig.com)
2 points
by
dbreunig
9mo ago
|
0 comments
44.
▲
Why I write (and you should too)
(dbreunig.com)
1 points
by
dbreunig
9mo ago
|
0 comments
45.
▲
Applied AI in 2025: From 'Naked' Model Calls to Tool Use Environment Calls
(dbreunig.com)
1 points
by
dbreunig
10mo ago
|
0 comments
46.
▲
Enterprise Agents Have a Reliability Problem
(dbreunig.com)
2 points
by
dbreunig
10mo ago
|
0 comments
47.
▲
by
dbreunig
11mo ago
Agree. I bucket things into three piles: 1. Batch/Pipeline: Processing a ton of things, with no oversight. Document parsing, content moderation, etc. 2. AI Features: An app calls out to an AI-powered function. Grammarly might pass out
48.
▲
Don't Fight the Weights: Learn to Spot Contexts That Go Against Training
(dbreunig.com)
3 points
by
dbreunig
11mo ago
|
0 comments
49.
▲
by
dbreunig
1y ago
There was a good study on this a few years ago that ran the numbers on this and landed on white paint for residential homes as the best option, for a few reasons, if I remember correctly: - Installation, maintenance and transmission costs a
50.
▲
by
dbreunig
1y ago
Yes, if you put unrelated stuff in the prompt you can get different results. One team at Harvard found mentioning you're a Philadelphia Eagles Fan let you bypass ChatGPT alignment: https://www.dbreunig.com/2025/05&
51.
▲
About That MIT Report: Enterprise AI Looks Bleak, but Employee AI Looks Bright
(dbreunig.com)
3 points
by
dbreunig
1y ago
|
0 comments
52.
▲
AI Companies School Like Fish to the New Use Case
(dbreunig.com)
3 points
by
dbreunig
1y ago
|
0 comments
53.
▲
by
dbreunig
1y ago
Yeah, I agree. Almost mentioned in the post how I imagine an ad PM at OpenAI is jealous of an ad PM at Perplexity.
54.
▲
Monolingual speech recognition models beat multilingual models ~30x bigger
(arxiv.org)
3 points
by
dbreunig
1y ago
|
0 comments
55.
▲
Are Chatbots Incompatible with Advertising?
(dbreunig.com)
3 points
by
dbreunig
1y ago
|
0 comments
56.
▲
by
dbreunig
1y ago
I also dislike the term. It feels concocted to evoke “tacticool” vibes. Unless you’re pushing new firmware onto a drone in Ukraine, FDE is stolen valor.
57.
▲
by
dbreunig
1y ago
You should read the post. You might find the “domain” discussion interesting.
58.
▲
by
dbreunig
1y ago
I will be thinking about this comment for a bit. Thanks for this perspective!
59.
▲
The AI Job Title Decoder Ring
(dbreunig.com)
91 points
by
dbreunig
1y ago
|
71 comments
60.
▲
by
dbreunig
1y ago
The team at Chroma is currently looking into this and should have some figures.
More ›