sosodev

11 ms·

181.
by sosodev 11mo ago
While that is part of the equation it's not at all that simple. If the average business owner wants a custom piece of software for their workflow how are they getting it now? For decades the answer would have been new hires, agencies,
182.
by sosodev 11mo ago
Ironically the practically of such instruction goes down as the status of the school goes up. I got a lot of 1:1 or 1:few time with my community college professors.
183.
by sosodev 11mo ago
But there’s no quantum explanation of gravity, right?
184.
by sosodev 11mo ago
The report doesn't say that because it's just reporting the facts not drawing conclusions. In my opinion, if a catastrophic failure happens that is a maintenance failure. Either the inspections failed to notice the fatigue or the
185.
by sosodev 11mo ago
I think Nano Banana Pro should have passed your giraffe test. It's not a great result but it is exactly what you asked for. It's no worse than Seedream's result imo.
186.
by sosodev 11mo ago
How do you figure? They're very similar planes. The left engine and its pylon detached in both cases during takeoff rotation. Both incident reports stated that proper maintenance would have prevented the detachment. The way the situati
187.
by sosodev 11mo ago
The MD-11 was developed after that crash. Shouldn't its design and maintenance procedures have been informed by the incident?
188.
by sosodev 11mo ago
A commenter in HN thread covering the initial crash mentioned that the left engine detaching might have been the cause https://news.ycombinator.com/item?id=45821537 The referenced AA Flight 191 is shockingly similar. It mak
189.
by sosodev 11mo ago
https://news.ycombinator.com/item?id=45890186
190.
by sosodev 11mo ago
Here's some insight from Jeff Dean and Noam Shazeer's interview with Dwarkesh Patel https://youtu.be/v0gjI__RyCY&t=7390 > When you think about divulging this information that has been helpful to your compet
191.
by sosodev 11mo ago
Wow, that's very impressive
192.
by sosodev 11mo ago
How can you be sure that your benchmark is meaningful and well designed? Is the only thing that prevents a benchmark from being meaningful publicity?
193.
by sosodev 11mo ago
I can't help but be reminded of the bitter lesson when I read about the continual spread of solar energy. The simple, scalable system wins over everything else. I wonder how many aspects of our lives could be transformed in this way.
194.
by sosodev 1y ago
Most AI companions have memory systems. It's still quite simplistic but it's not "goldfish memory" or just limited to the context window. From what I understand, some inputs from the user will trigger a tool call that se
195.
by sosodev 1y ago
That's just semantics. The $200 is the deposit. It's $20,000 to actually buy it once they're ready to ship.
196.
by sosodev 1y ago
Does it support OpenRouter? I tried configuring OpenRouter as a "local model" but it seems to silently fail.
197.
by sosodev 1y ago
LLMs don't "read" text sequentially, right?
198.
by sosodev 1y ago
This article makes the claim that OpenAI, and AI in general, is massively overhyped because OpenAI is looking to sell slop. I'm not sure I can agree with that basic premise. Whether AGI does or does not materialize sometime soon doesn&
199.
by sosodev 1y ago
Thank you. Reflexive vs reflective is a good way to frame it. You're right that the reflective comments do show up eventually so I suppose I need to be more patient too.
200.
by sosodev 1y ago
I think it's a shame that a 146 minute podcast released ~55 minutes ago has so much discussion. Everybody here is clearly just reacting to the title with their own biases. I know it's against the guidelines to discuss the state of
201.
by sosodev 1y ago
I swear HN commenters say this about every frontier model.
202.
by sosodev 1y ago
It's easier to develop a game for a mac in some ways but you reach a tiny fraction of gamers that way.
203.
by sosodev 1y ago
I suspect that we will see it when an OTA update very rapidly causes hundreds of deaths.
204.
by sosodev 1y ago
Isn't that assuming training methods remain the same? It seems like learning from the environment will be a requirement for robots to scale. My understanding is that research has been yielding new architectures that might have that typ
205.
by sosodev 1y ago
So breaking the law is ok if you don't agree with it?
206.
by sosodev 1y ago
CS2 is mostly 64 tick from what I understand. The "sub-tick" stuff is timestamping actions that happen on a local frame before the next tick. So in theory the client feels perfectly responsive and the server can adjust for the del
207.
by sosodev 1y ago
Flagging is definitely getting abused more on HN lately. The consensus seems to be that politics and/or morals are irrelevant outside of personal affairs so we must not have these conversations here.
208.
by sosodev 1y ago
Your friend's issue isn't whether he is or isn't autistic. It's that he gave up his agency.
209.
by sosodev 1y ago
The bot will only do whatever you tell it if that's what it was trained to do. The same thing broadly applies to humans. The topic of free will is debated among philosophers. There is no proof that it does or doesn't exist.
210.
by sosodev 1y ago
Their description would be bad without some prior training of course but so would the LLM's.

More ›