asdev

6 ms·

211.
by asdev 2y ago
do you have an eval? how did this actually help?
212.
by asdev 2y ago
Absolutely not. Nothing like an out of touch manager slowing down their engineers with random PR comments and questions. They should not go deeper than the design document/architecture level. They are not adding any value trying to put
213.
by asdev 2y ago
why would use this when I can just add API docs to my LLM context and have it generate the integration code?
214.
Ask HN: Why did you start your startup?
2 points by asdev 2y ago|0 comments
215.
by asdev 2y ago
if you cooked a meal with a ton of butter or oil, it would not be able to detect the quantity. it could possibly undercount by hundreds of calories, making it useless. if you're not serious about the counting then it could be fine
216.
by asdev 2y ago
this idea has been rinsed and repeated many times already. unless you're going to include manual input in addition to the image recognition, it will be highly inaccurate
217.
by asdev 2y ago
CalAI is a scam. It's highly inaccurate, there's no way to determine calories from a picture for a multitude of reasons
218.
by asdev 2y ago
the majority of software is just a bunch of rats nest business logic and CRUD DB ops. hence it's a CRUD app. Queues, microservices etc are all over engineering for no reason
219.
by asdev 2y ago
did you have something to uniquely differentiate your alternative? you don't need to reveal what it is, just wondering if you noticed a gap to fill
220.
by asdev 2y ago
that means the lean startup method is dead because if you need to come out of the gate with a polished product, you need to build for a while before launching
221.
by asdev 2y ago
yeah but I can't build anything for them without the domain expertise in German immigration or some other domain. too far removed from the problem
222.
Ask HN: Is SaaS Dead?
26 points by asdev 2y ago|41 comments
223.
by asdev 2y ago
who has productionized an agent in a setting where there is a low margin for error? I would love to know
224.
by asdev 2y ago
best Leetcoders generally add the least value business wise
225.
by asdev 2y ago
if you write your code with good dependency injection/abstraction, you can one shot unit tests a lot of the time
226.
by asdev 2y ago
TLDR retention is more important than top of funnel
227.
by asdev 2y ago
Companies are still not being penalized for spending exorbitant amounts of money gambling on AI "scaling"
228.
2025 will likely be another brutal year of failed startups, data suggests(techcrunch.com)
4 points by asdev 2y ago|0 comments
229.
by asdev 2y ago
Are we finally realizing the "AI Agents" are complete snake oil? Cron jobs and while loops with non deterministic crappy behavior?
230.
Ask HN: Is anyone using Function Calling with LLMs in production?
5 points by asdev 2y ago|1 comments
231.
by asdev 2y ago
Perplexity is going to be a nice headstone in the AI graveyard once the bubble pops
232.
by asdev 2y ago
I don't know any half decent coders who use Replit.
233.
by asdev 2y ago
how much more confirmation do we need that we are in a recession?
234.
by asdev 2y ago
The main point here seems to be around boxing in the AI with a rigid set of rules, and then not being able to adapt to model improvements over time. I think if those "rules" are just in the prompt, you can still adapt, but if you&
235.
by asdev 2y ago
an ad machine*
236.
by asdev 2y ago
I hope the stock gets destroyed next earnings when people realize they've been spending all this money on AI all to not ship a single meaningful feature nor derive any revenue leveraging it. They did the exact same thing with the whole
237.
by asdev 2y ago
it was never an economic boom, just an inflationary boom. AI hype created an echo bubble from 2021 but it's due to pop
238.
by asdev 2y ago
AI generated landing pages are quite obvious
239.
by asdev 2y ago
did you market it at all? or people just found it organically
240.
by asdev 2y ago
I've tried both Copilot and JetBrains AI with IntelliJ and both are awful compared to Cursor. No multiline editing, no composer, worse at writing tests etc.

More ›