a h

483 posts

a h banner
a h

a h

@x97115104

builder & privacy advocate | emerging tech enthusiast; views are my own. -- also https://t.co/9xyOcEHWtb 🦋

Katılım Kasım 2024
420 Takip Edilen80 Takipçiler
Sabitlenmiş Tweet
a h
a h@x97115104·
Stop frontier providers from eating your lunch. Own your AI bill by switching to open source!
English
1
0
4
373
a h
a h@x97115104·
The difference between how a company treats us when they make a mistake can be the ultimate in loyalty building or destroying. A mistake handled well might make us more loyal customers than we would have been had we never experienced a service problem.
English
0
0
0
11
daniel wintermeyer
daniel wintermeyer@danwintermeyer·
we're hiring btw for 2000 roles. if you're in engineering or gtm, hmu.
English
143
15
547
82.4K
a h
a h@x97115104·
The difference between how a company treats us when they make a mistake can be the ultimate in loyalty building or destroying. A mistake handled well might make us more loyal customers than we would have been had we never experienced a service problem.
English
0
0
0
29
a h
a h@x97115104·
It is least effective when asked to replace architectural judgment, domain expertise, or responsibility.
English
0
0
0
23
a h
a h@x97115104·
Agentic AI is most effective when it accelerates understanding, validation, and repetition.
English
1
0
0
28
a h retweetledi
Paul Graham
Paul Graham@paulg·
A watchmaker on watch winders: "Don't use autowinders. They grind your watch to dust."
English
59
9
446
127.7K
a h
a h@x97115104·
@paulg yes, let’s make said compiler. we’d just have to get around bayesian probabilism in terms of deterministicly gauging such code, most “ai detectors” are unsuccessful for this reason. attestations/watermarking at the generation level might work tho! interesting idea.
English
0
0
3
691
a h retweetledi
Paul Graham
Paul Graham@paulg·
College students use AI to do most of their writing. An increasing number of professors secretly use it for grading. In the limit case, AIs do all the work, and all the humans do is transmit what they create. A good compiler would recognize this as dead code and remove it.
English
383
354
6.4K
388.1K
Megan Lentz
Megan Lentz@megmlentz·
people keep telling me X has the most cracked engineers on the internet.. let's test the theory. @usecorgi is hiring. drop me your GitHub👇
English
43
8
57
6.3K
a h retweetledi
Eric Lay
Eric Lay@itsericlay·
I give up. Here are all the eng. roles we're hiring for: - Founding Design Eng. $150K - $250K - Founding Harness Eng. $130K - $200K - Founding Infra Eng. $160K - $250K The comp is all cash base salary and you get significant equity ($400K+) + we offer relocation and visa support. Please help and support me by: 1. reacting, commenting, reposting 2. share with anyone who might be interested in joining a fast growing startup 3. apply directly with the link in the comments 4. pray @doubleemt will extend the time to hire
English
59
31
474
83.9K
a h retweetledi
François Chollet
François Chollet@fchollet·
The true measure of a software engineer isn't their ability to write clever code. It's their ability to ruthlessly protect the codebase from unnecessary cleverness.
English
88
193
2.2K
82.4K
Deborah Bentham 🏳️‍⚧️🏳️‍🌈
@x97115104 @RaminNasibov My dad & his boss got thrown out of a Dale Carnegie course. They spent the whole time pointing out the shortcomings of the methods & they were asked to leave as they were making some of the course leaders’ question the principles. 😂😂😂
English
1
0
1
47
Ramin Nasibov
Ramin Nasibov@RaminNasibov·
What’s a book everyone hypes up but is actually total trash ?
English
213
10
93
442K
a h retweetledi
steven
steven@syriansigma·
We're paying $150–250k base + equity for AI engineers at Weave, and I'm not reading a single resume. Instead: 1. Sign up free → workweave(dot)ai 2. See your code output data 3. Post a screenshot If you're cracked, I'll reach out directly. Takes 2 minutes, and your code does the talking.
English
88
16
571
494.7K
a h retweetledi
Reads with Ravi
Reads with Ravi@readswithravi·
Be delusional about your potential. Be delusional about your potential. Be delusional about your potential. Be delusional about your potential. Be delusional about your potential. Be delusional about your potential. Be delusional about your potential.
Reads with Ravi tweet media
English
52
2.2K
12.4K
177.4K
a h
a h@x97115104·
what would models paying for their own inference look like? how might they pay for their own GPU hours? 🤔
English
0
0
1
32
a h
a h@x97115104·
it's not really how smart you are or how hard you work or how efficiently you work. the work you decide is valuable is what determines your outsized impact, so choose well!
English
0
0
0
23
a h
a h@x97115104·
the longest horizon task is deciding for yourself what should be done, and then doing it. deciding what to do is an unbounded length task, so where you spend your time is what determines how successful you are, and by extension, how successful your agents will be.
English
1
0
0
29
a h
a h@x97115104·
what are you working on?
English
1
0
0
35