mplewis

11 ms·

211.
by mplewis 2y ago
Please don’t spread ivermectin conspiracy theories on HN.
212.
by mplewis 2y ago
Buddy, if you can’t charge your mouse once after weeks of notifications, that’s a you problem.
213.
by mplewis 2y ago
Two orders of magnitude would be 100, not 10.
214.
by mplewis 2y ago
Yeah, I feel so sorry for the men who lose their status when they leave behind patriarchy. Won't someone think of the poor men. :eyeroll:
215.
by mplewis 2y ago
I need more information on what this can do before I'm willing to create an account.
216.
by mplewis 2y ago
From my experience, I predict that LLM code completion will not work very well when using a language like Civet. Copilot is good when working with something where the idiomatic patterns are common, widespread, and hard to screw up. Civet ha
217.
by mplewis 2y ago
In the USA, the TSA fails to detect threats 95% of the time. "Not confident" is an understatement: we know that these procedures do not work.
218.
by mplewis 2y ago
How many DAU do you think Hacker News has?
219.
by mplewis 2y ago
There was never a time in human history when Senior Software Engineer had a consistent definition. It’s fine to write a nostalgia piece about your past experiences, but this blog post doesn’t go very far beyond that.
220.
by mplewis 2y ago
useState has been present in React for years.
221.
by mplewis 2y ago
That’s incorrect. See: https://en.wikipedia.org/wiki/Master_gardener_program
222.
by mplewis 2y ago
How does your project compare with the completely free-to-use citizen science project and mobile app, PlantNet? https://plantnet.org/en/about/
223.
by mplewis 2y ago
I’m betting on the collapse of “LLM as replacement for human labor” because it fundamentally doesn’t solve that problem.
224.
by mplewis 2y ago
Wow, this thing can reason now? How come it keeps getting my basic word problems wrong?
225.
by mplewis 2y ago
Tells you a lot about the state of the industry.
226.
by mplewis 2y ago
There is no heuristic that reliably separates human-written text from GPT text. It is impossible to avoid model collapse in such an environment.
227.
by mplewis 2y ago
It’s literally the same product as all the other LLM competitors. If you kill ChatGPT, users will be on Claude in about three seconds.
228.
by mplewis 2y ago
A few generations of what, pretending their robots work?
229.
by mplewis 2y ago
How much do you think it should cost?
230.
by mplewis 2y ago
Automattic needing more money explains all of Matt M's moves.
231.
by mplewis 2y ago
As described in the other thread, this is known as the inferential mood in Turkish: https://en.wikipedia.org/wiki/Inferential_mood
232.
by mplewis 2y ago
The inferential mood (your "gossip tense") is more related to mood (signals a particular relationship to truth or reality) than tense (signifies a relationship to time).
233.
by mplewis 2y ago
They were already contributing as evidenced by the existence of ACF.
234.
by mplewis 2y ago
80% of homelessness in the USA is transitional, i.e. someone lost a job or had an injury and needs time to get back on their feet.
235.
by mplewis 2y ago
Cars that drive in a circle around a fixed course? Yeah, I've seen that before.
236.
by mplewis 2y ago
It's 2024 and the Optimus can almost do some of the things Asimo could in 2000.
237.
by mplewis 2y ago
Wow, great! They put motors and wheels on a chassis with a screen and hardcoded them to drive in a loop. So impressive.
238.
by mplewis 2y ago
He just bought a massive data center with thousands of Nvidia boards. Why would the car be an efficient place to do the same work?
239.
by mplewis 2y ago
Because all the work of allegedly building a new vehicle platform is a better excuse for making no profit on the robotaxi initiative for the next ten years.
240.
by mplewis 2y ago
Rust isn’t popular in web dev. It’s very popular in embedded.

More ›