would be interesting to ran an analysis that shows overlapping text with existing songs, as it just sounds like bits of everything smooshed together
what happens when it hallucinates? do you end up with the machine probing in your mouth or eye sockets instead?
Somewhat Ironically as a Linux user I have glasses and full beard with messy curly hair, fuck I'm a flipping stereotype.
I guess I'll need to go skinhead and drop the glasses, hit the gym and pump some iron.
This quoted post is unavailable.
"Hard times create strong langs, strong lang create good times, good times create weak langs and weak langs create hard times."
Yep fits perfectly...
"find the problem" assumes you fully understand the original problem at hand, because without that one mans problem might be another mans solution? Sure might work for cute little puzzles maybe, not so much real world codebases that are even barely understood by humans
High income with higher taxes and even higher living costs and lowest output in terms of actual value and satisfaction = miserable lifestyle
๐ฌ๐ง British people are now the 2nd most miserable people in the world
Previous happiness studies are based on income, wealth and healthcare quality, simply putting Western countries on top always, very very biased
The new Mental State of the World report is the first happiness ranking that actually makes sense and is based on surveys of satisfaction
People in many low-income countries are actually MORE happy regardless of income: in this case for example ๐ฉ๐ด Dominican Republic and ๐ฑ๐ฐ Sri Lanka are VERY happy
While people in a high-income country like United Kingdom are VERY unhappy
Added the data to Nomad List for the [ โบ๏ธ Happy ] filter
and how exactly would that work? If the first statistical model itself can't disambiguate between sense and nonsense, how would a second statistical model on top of first solve that? It took would suffer the same recursive issue...
He's my spicy take:
Companies jumping on the AI bandwagon now in a few years will generate a mountain of "AI Shitcode", this shit will need maintaining/fixing, for that will require actual devs, the AI hype train will crash and burn, and the tug of war for real devs will ignite.
LLMs are garbage, I've seen the shit code it generates, half the time it hallucinates utter nonsense while the other half I could have searched that shit from GitHub where it most likely come from in the first place. Other then a better search engine, what has LLMs really done?
So basically this looks suspiciously like those CLI wizards that bootstrap a project. Still unconvinced that this is any better than a framework. And what about when it hallucinates... ๐
Wow a new language "Dada" like Rust but with small runtime ๐คฉ
Yeah I'm up for that shit!
Wait gradually typed? ๐คฎ
Close browser immediately.
dada-lang.org/