4 ms·

Paul Buchheit: We all have tunnel vision

43 points by toffer 19y ago

Harj 19y ago
one way to avoid this is deliberately trying to find non-corroborative evidence for every idea/opinion you form. it's talked about in the Black Swan and is surprisingly difficult to do.
- aston 19y ago
  I actually have no idea what "non-corroborative evidence for an idea" means. Does that mean try to prove yourself wrong? Or to try to find evidence for the idea that, due to its difference from your original evidence, doesn't actually corroborate? Or ...?
  - aswanson 19y ago
    Just give audience to, or work very hard to find contrary evidence for, a concept you hold as true.
  - pc 19y ago
    "I actually have no idea what "non-corroborative evidence for an idea" means." It's the opposite of http://en.wikipedia.org/wiki/Corroborating_evidence http://en.wikipedia.org/wiki/Corroborating_evidence
    - tokipin 19y ago
      not to be technical, but i believe "non" means "absense of," not "inverse of." so for a given statement there's a lot of non-corroborating evidence that is irrelevant
henning 19y ago
Suggestion: stay the fuck away from reddit, at least anything besides the science and programming subreddits. Even frequenting programming.reddit you might get the idea that functional programming is poised to break into the mainstream any day now. Don't hold your breath.
- SwellJoe 19y ago
  "Even frequenting programming.reddit you might get the idea that functional programming is poised to break into the mainstream any day now. Don't hold your breath." I think you're wrong on this one. In the Perl community (a mainstream community if ever there was one), Haskell is very popular, and a rather large swath of interesting work around Perl is in functional techniques and tools. Functional programming is definitely going mainstream. Not to mention that Simon Peyton-Jones works for Microsoft and F# is getting quite a push.
  - davidw 19y ago
    My bet is to see someone "steal" a lot of FP ideas and integrate them into something aimed at being a more mainstream language, rather than Haskell or something of its ilk becoming popular. That seems to be the trend with things like Lisp and Smalltalk.
- andyn 19y ago
  I agree with on the reddit point, it seems to have fallen into the trap of each article vying for attention and becoming a pit of exaggerated headlines toeing the party line. Run away! I don't think that functional programming languages are about to break into mainstream despite the gushing reviews out there. I think it's more akin to learning Latin as a language - I found Lisp very interesting and I think my programming has improved as a result of learning it but I don't think I'll ever use it in a real-life project.
downer 19y ago
> Supposedly people can hold about seven "items" in their mind at any one time. I was never sure what that meant -- what qualifies as an "item"? This urban legend PB refers to actually originated with studies of short-term memory and remembering sequences of digits. See http://72.14.253.104/search?q=cache:PP_hzY7PqKgJ:www.knosof.co.uk/cbook/misart.pdf http://72.14.253.104/search?q=cache:PP_hzY7PqKgJ:www.knosof....
myoung8 19y ago
I think YC gives us tunnel vision too... Maybe it's not such a bad thing, though.

43 points by toffer 19y ago

Harj 19y ago

one way to avoid this is deliberately trying to find non-corroborative evidence for every idea/opinion you form. it's talked about in the Black Swan and is surprisingly difficult to do.

aston 19y ago
I actually have no idea what "non-corroborative evidence for an idea" means. Does that mean try to prove yourself wrong? Or to try to find evidence for the idea that, due to its difference from your original evidence, doesn't actually corroborate? Or ...?
- aswanson 19y ago
  Just give audience to, or work very hard to find contrary evidence for, a concept you hold as true.
- pc 19y ago
  "I actually have no idea what "non-corroborative evidence for an idea" means." It's the opposite of http://en.wikipedia.org/wiki/Corroborating_evidence http://en.wikipedia.org/wiki/Corroborating_evidence
  - tokipin 19y ago
    not to be technical, but i believe "non" means "absense of," not "inverse of." so for a given statement there's a lot of non-corroborating evidence that is irrelevant

henning 19y ago

Suggestion: stay the fuck away from reddit, at least anything besides the science and programming subreddits. Even frequenting programming.reddit you might get the idea that functional programming is poised to break into the mainstream any day now. Don't hold your breath.

SwellJoe 19y ago
"Even frequenting programming.reddit you might get the idea that functional programming is poised to break into the mainstream any day now. Don't hold your breath." I think you're wrong on this one. In the Perl community (a mainstream community if ever there was one), Haskell is very popular, and a rather large swath of interesting work around Perl is in functional techniques and tools. Functional programming is definitely going mainstream. Not to mention that Simon Peyton-Jones works for Microsoft and F# is getting quite a push.
- davidw 19y ago
  My bet is to see someone "steal" a lot of FP ideas and integrate them into something aimed at being a more mainstream language, rather than Haskell or something of its ilk becoming popular. That seems to be the trend with things like Lisp and Smalltalk.
andyn 19y ago
I agree with on the reddit point, it seems to have fallen into the trap of each article vying for attention and becoming a pit of exaggerated headlines toeing the party line. Run away! I don't think that functional programming languages are about to break into mainstream despite the gushing reviews out there. I think it's more akin to learning Latin as a language - I found Lisp very interesting and I think my programming has improved as a result of learning it but I don't think I'll ever use it in a real-life project.

downer 19y ago

> Supposedly people can hold about seven "items" in their mind at any one time. I was never sure what that meant -- what qualifies as an "item"? This urban legend PB refers to actually originated with studies of short-term memory and remembering sequences of digits. See http://72.14.253.104/search?q=cache:PP_hzY7PqKgJ:www.knosof.co.uk/cbook/misart.pdf http://72.14.253.104/search?q=cache:PP_hzY7PqKgJ:www.knosof....

myoung8 19y ago

I think YC gives us tunnel vision too... Maybe it's not such a bad thing, though.