I think it's less that politicians think short term, and more that voters think short term. Politicians are just responding to the incentives created by willfully ignorant voters.
It's amazing that California has so many horrible problems with housing shortages, poverty, and basic governance... and yet this is what representatives choose to concoct.
AirBnBs are not responsible for the housing crisis. Voters that vote to block development at the municipal level are responsible for the housing crisis.
Maps & Gmail & Search all have plenty of accumulating flaws... but they also completely defined their product category and today are among the most popular software products ever made.
If you're a materialist and you think there is no afterlife then I don't believe there would be a void. There's no mechanism for "you" to perceive a void or perceive time.
Its the difference between:
"I see nothing when I close my eyes"
(actually you 'see' darkness over time)
and...
"I see nothing through my elbow"
(You really don't see anything through your elbow because the capacity for it just doesn't exist.)
Even our basic perception of time is mediated by the structure of our brains... and without that there's no difference between one second and a trillion^trillion years.
I mean the goal is that the money invested increases the value of the company so that the founder's reduced stake is worth considerably more than the founder's full stake of the counterfactual company where he didn't take VC money. VCs can be strongly positive sum.
Because LLMs are not humans, and the code they produce will have a different distribution of failure modes than human written code, so attribution is useful info while reviewing?
It's fascinating that Clojure has consistently the best performing solutions and yet at the same time such a low success rate.
Do you have an idea as to why that is?
If I had to guess, two things lowering reliability:
A) Balancing parens might be tough on an LLM one-shot.
B) LLMs generate tokens sequentially, but s-expressions mean the first forms to be evaluated in a body are usually the last to be written, so the LLM has to sequentially generate layers of evaluation backwards.
As someone who loves Clojure, I wonder about the real portability across host languages. Do you have experience with any of these other dialects? (beyond the obvious CLJS & Babashka?)
Absolutely agree. Recently learning emacs with vertico/consult/orderless/embark has had me thinking constantly about how needlessly crippled most OS windowing system workflows are.
It is absolutely painful for interviewers and candidates. I used to be able to email managers & founders directly and discuss what they were looking for... today everybody is navigating a deluge of spam and the interview process is becoming dysfunctional.
I agree. I'm as skeptical as many commenters but I also think the degree of polarization in HN around this technology and the degree to which people are calling those with different views shills or naysayers is pretty sad.
There's nothing sneaky about terms & conditions. If the gov wants a service they legally need to abide by its terms, same as us, if they don't like it they should choose another product.
Anthropic doesn't want their AI used for misaligned mass surveillance scanners and killbots, there are obvious reasons they might not want that.
I'm sure you're right that AI augmented workflows can (& do?) produce beautiful works that I would call art... it's just that the overwhelming majority of AI 'art' I experience on the internet is slop.
There are so many basic gaps in functionality and so many underbaked & poorly designed Mac OS features that I end up papering over with paid 3rd party applications.