don't have time to write up my favorite books of 2015, so here's the whole shelf worrydream.com/Shelf2015
Four years ago I was leaving their house into the snow and Gloria said "you need a hat" and gave me Marvin's hat.
"As we have never been cast down with scorn and ridicule we shall never be puffed up with praise and admiration" -ECS
Somehow, returning to caltech regresses me to an awestruck 18-year-old.
I opened that nightstand drawer where hotels usually put a bible, thought maybe I'd find the feynman lectures
i feel pretty concerned tbh
potential sapiens
@michael_nielsen: The answer to this question ["Can computers be creative?"] seems so obv "yes" to me, I can't imagine otherwise. And yet others strongly disagree.
@michael_nielsen Isn't a better question "SHOULD computers be creative"? ie: Why are humans creative? Is outsourcing that good for humanity?
@michael_nielsen How do humans benefit from other humans being creative? Do we get that benefit from creative computers? Or miss the point?
@michael_nielsen There's an argument parallel to what @zephoria said about social media:
Beautiful work from @unconed. Imagine these visuals not as "explainers" but just "the stuff you work with every day"
acko.net/files/gltalks/…
Life milestone achieved: trapping Scott McCloud's family and pets inside a dark locked freight elevator
I don't know.
UPDATE: I still don't know.
We shall not think of thinking, without imagining ourselves thinking *somewhere*.
archive.org/details/timelessway…
The thought and the space shall be indivisible. The thought is supported by the space. The space supports the thought.
smart posters, not smart toasters
Developers really need to make a medium that doesn't cripple people into being helplessly dependent on "developers".
Developers really need to make a medium that offers people at least as much power and freedom as a piece of paper and a Sharpie.
I wonder how much Strong AI research is ultimately driven by being lonely and wanting a friend to talk to
@mayli: While
@deray
taken by cops as he is streaming / reporting protest, his Twitter stream has animated balloons
If a human newspaper editor juxtaposed this news with party balloons,
the outrage would force them to resign.
But when an algorithm does something grossly inappropriate in a
somber moment, we just accept it. "Poor algorithm doesn't know any
better."
Designers, please think about this before adding that "fun" or "playful" feature or
"attitude". Your app doesn't understand context.
@scottjenson: I appreciate your point but the only practical alternative is to never celebrate birthdays this way. Baby with the bath
Really? People have been celebrating birthdays for a long time
without algorithms reminding them.
If it's the thought that counts, then a database query triggering a
mindless animation is the opposite of a "celebration".
@scottjenson: So no website/app can ever say 'happy birthday' ever again? That's a bit strong isn't it?
The entire point of birthdays is human bonding. A robot saying
"happy birthday" is entirely meaningless.
@mayli: also of note: automated regular flower delivery to a loved one is *not* romantic. And intention != effect
@nsfmc: weird concept tho—where do you draw the line if the observed
effect is identical? 🤖💐? cal ⏰? string on ☝🏽? memory? moleskine?
Effect isn't identical. Flowers sent by human means "I'm thinking
of you". Flowers sent by robot means "a robot is thinking of
@nsfmc: i was using effect as what recipient sees "flowers arrive".
agree intent can vary wildly—decoding intent is complex
I think of the effects as "recipient (correctly) knows someone is
thinking of them" or "recipient is deceived" respectively.
People aren't black boxes. :)
"Amusing Ourselves to Death" is not an instruction manual.
"So can you talk to me about how MacPaint started? Was there a concept at design meetings?"
"Hahahaha no no no no."
youtube.com/watch?v=-syl7m…
"Hahahaha no no no no."
youtube.com/watch?v=-syl7m…
In honor of the passing of Jerome Bruner, why don't you stop recapitulating behaviorism using neural networks.
symbolics probably failed because genera didn't have a "cloud family restrictions daemon"
@DanielleFong: What if, at the "age of consent" they asked you, "do you consent to society?" What happens if you say no?
the entire world is full of magic
Kind of refreshing to see marketing copy that emphasizes patience and learning curve instead of instant expertise.
(Ironically, it seems like Bluespec is actually pretty easy to learn, and extends naturally from Verilog?) youtube.com/watch?v=YpiLEo…
@michael_nielsen: An image that implements FizzBuzz, done in Piet, the image-based programming language
@michael_nielsen An image that implements an entire processor, done in the original image-based programming language
I've reached the point where I assume that any citation in any written work is, as likely as not, a misunderstanding.
(probably going to delete this tweet because I'm pretty sure it'll be misunderstood)
If you want a vision of the future, imagine a million tweets talking past one other, forever.
"Grandpa, why did Hazel-rah die?"
"Back then, everybody grew old and died, even the rich and powerful."
"Grandpa, it's hot and I'm hungry."
"Back then, everybody grew old and died, even the rich and powerful."
"Grandpa, it's hot and I'm hungry."
The typesetter wipes his brow with his sleeve, takes a swig of whisky, unlocks a box labeled THE BIG INTEGRAL.
Carver Mead has a new website, and it has everything.
carvermead.caltech.edu
Really nice introduction to Modelica by @mtiller, including a demo of a literal "npm for scientific models."
youtube.com/watch?v=-mvEUu…
youtube.com/watch?v=-mvEUu…
Modelica is such a fascinating in-practice example of truly declarative components that have no "inputs" or "outputs", just relationships.
Hi. Does this give you feelings?
complex.com/life/2016/09/e…
You know what you need to do.
worrydream.com/ClimateChange
complex.com/life/2016/09/e…
You know what you need to do.
worrydream.com/ClimateChange
Don't panic. Don't despair. Build.
If you call yourself an engineer, it's time to step up and start engineering.
worrydream.com/ClimateChange
If you call yourself an engineer, it's time to step up and start engineering.
worrydream.com/ClimateChange
Neither "Glass" nor "Spectacles" will enable people to see the way out of global crises. Stop bullshitting. Build what the world needs.
@paulg: Do you mean that only working directly on problem x will help solve problem x?
@paulg The second half of the essay is entirely about high-leverage projects that indirectly help solve problem X. worrydream.com/ClimateChange
@paulg High-leverage indirect is extremely important. But it's unlikely that addressing lifestyle inconveniences will turn out to be relevant.
@DynamicWebPaige: "Gosh, I really wish I would have majored in mathematics and computer science." ←my brain, at least twice a week for the last four years
@DynamicWebPaige Never regret studying how the real physical world works, instead of some abstract self-referential fictional hocus-pocus.
@DynamicWebPaige: It'd be difficult to learn how the real physical world works, though, without having a set of observational/explorational tools
@DynamicWebPaige Of course, and studying physical science and engineering is all about learning and developing those mathematical tools.
@DynamicWebPaige But unlike CS, you can be sure those tools refer to something outside themselves. CS can be all maps, no territory.
@DynamicWebPaige I'm not saying CS techniques aren't useful. But if *all you know* is CS, all you know is fictional worlds,
@DynamicWebPaige and your tendency is to hide in them and drag others into them, instead of knowing how to deal with reality.
Reading about the clumsy ad-hoc compression method used by HP's
Smalltalk implementation when writing the object table to an image...
... I thought, "why didn't they just LZ77 it", and realized with a shock
that this wasn't long after 1977, and LZ77 wasn't widely known.
... I can't even imagine what it would be like to not have LZ77. It's like
imagining life before the discovery of fire.
@posco: It's a real loss that modern society is so age segregated. Children remind us to play and what joy is. And elders have much to teach.
Figure 9. The continuum between "things" and "stuff". (@dSeanMustard ) dl.acm.org/citation.cfm?i…
it's so far away
with every step it feels further away
with every step it feels further away
hm, I think "join a mob" would have been better for the twitter line, oh well
software is eating our words
almost exactly eight years from concept to production
twitter.com/worrydream/sta…
tbh, makes me feel better that I'm not planning to show you what I'm currently working on for eight years!
a blue plane in pink clothing
"What must we do to make 'possibly' into 'probably' in two years?" (from amazon.com/dp/0143122797)
(via @patrickc) continuing to work on long-term research is a kind of Pascal's Wager
@ncasenmare:
For others reading this: w/ Score Voting, it's possible for someone who has a majority to NOT win ncase.me/ballot/sandbox...
My favorite part of Nicky's latest is this twitter reply which links a custom interactive simulation to illustrate the point.
Nicky's article is excellent (as always!), but to me, the deepest use of "explorable explanations" is not as standalone "explainers",
but as *conversational* tools, so everyone involved in a discussion can see, explore, and challenge what's being discussed.
The explainer can be thought of as introducing/teaching a language/toolset/interface/representation that can then be used in discussion.
What's your "fundamental theorem"?
@mayli: what's yours?
100% agree with @MaxCRoser about limitations of stories. But stats are very limiting too (e.g. Seeing Like A State)
ourworldindata.org/a-history-of-g…
Stories & stats are two extremes, both myopic: one too specific, one too general. Neither can represent the truth of a complex situation.
But stories & stats are relics of the paper medium. The dynamic medium allows a "ladder of abstraction" between them.
worrydream.com/LadderOfAbstra…
I believe that the deepest understandings will come from a new form of "reading" which is actively moving up and down such a ladder.
@wynlim: yet how do we nurture more people capable of creating such content/formats?
100 years of inventors messing around; followed by a @EdwardTufte/Euclid-like consolidator/teacher; then curriculum + community of practice.
Step 1 requires a *medium* that makes it possible for many people to mess around and learn from each other. For infographics, that was print.
montereybayaquarium.org >> "Conservation & Science" >> "Publications" >> "Sea Otter Publications" >> [list of 140 papers on sea otters]
Seymour Papert, Jerome Bruner, Marvin Minsky, and Jay Forrester died this year.
Seymour Papert (skip to 10:20): c-span.org/video/?67583-1…
Jerome Bruner: youtube.com/watch?v=xxn6Ip…
Marvin Minsky: youtube.com/watch?v=EI0NXT…
Jay Forrester: youtube.com/watch?v=JZLpbh…