Paul Graham Keynote at PyCon
brianrue.wordpress.com
brianrue.wordpress.com
I see these categories of email
1. Email from people you know that you're likely to respond to, and which therefore lead to "conversations"
2. Reminders and "things to do" that you need to act upon; once you have acted upon it you usually update the status of the task by mailing back
3. Read-only email: email sent by corporations that you somehow need to store but that you will never respond to (they often come from aptly-named "no reply" addresses; but mailing lists are part of it too)
4. Random stuff (close to spam, but not really): things you have to read because they may be important, but usually aren't
Gmail is optimized for #1 and does a good job at it. Lately it started to address #4 by trying to automatically detect "important" email; in my opinion this doesn't work well, but it could eventually work.
AFAIK nobody is seriously trying to solve #2 and #3, and in my experience #3 and #4 cause most of the clutter. Which makes #3 the most urgent problem with the highest ROI.
What if you could use a special email address for "read-only email" that would receive #3 mail and deal with it:
- organize it automatically by sender (or "topic") and period (not by subject / conversation / date: all my utility bills of one year are ONE object and not a dozen different conversations, for example)
- save all attachments (that deserve it) as PDFs
- notify you only when necessary (not easy)
- etc.
I've been toying with mailgun today, and it seems building a first version of this isn't out of reach; I'd be very interested to exchange ideas on this subject.
Why is this the case? Probably because we've only been able to get this firehouse of information for a relatively short period of time, so it's new and interesting.
I think the bigger opportunity is to step back and think about what our homepages will look like in 50 years. Somehow I doubt it will have anything to do with finding basic facts. Finding a better way to organize spam is thinking too small, I think you need to figure out a better way to integrate conversations and information into our broader life goals and daily lives.
If you have your own domain (which I think most people have on HN?) you can use different usernames to the same effect, or a different separator.
But even if it worked all the time it's only part of the solution.
* Notification emails that are very time-specific in their relevance. "Reminder: Coffee with Chuck at 3:25 today" or "Thank you for your reservation for a party of 4 at Alice's Restaurant at 7:30 tonight." You care about these a lot, until you have coffee with Chuck and abruptly stop caring.
* Things that are nice to have pass through your consciousness at some point. Amazon shipping confirmation emails, GitHub comment notifications, etc. Interestingly, most of these are mostly noise; when I get an Amazon confirmation email, the part I care about could fit in a Tweet: "Your Dromiceiomimus Plushie has shipped via USPS, with tracking number 54989086, and will arrive between Tuesday and Friday".
* Mailing lists. I have Gmail set up to tag all these and skip my inbox, which seems to be about right. The weird thing about mailing list emails is that almost all of them are explicitly marked as such -- there are several unofficial MIME headers for this -- and yet the default behavior for every email client I know if is to simply stick them in the inbox like any other email.
* Crap I don't care about. "Save up to 25% on useless ornamental doilies!!!1". The only part of this that I care about is the Unsubscribe link. I believe Gmail can auto-unsubscribe you if you mark these as spam.
Of course, while it's fine to break it down like this, the trouble starts when you start trying to actually extract useful information from emails so you can present users with a better view of their inbox. It's officially a Hard Problem, though like a lot of hard problems, you can get something useful with a lot less effort than getting something perfect.
What would you like from your inbox? For obvious reasons, I'm very interested. :-)
Your other examples are clearly #3... and I don't need to read any of those emails.
Amazon confirmation emails are completely useless since Amazon always ships on time; in the exceptional case where a package doesn't arrive, I can go fetch the tracking number somewhere to investigate; but reading these for every package is a waste of time.
Mailing lists should go to a special folder labelled "things to read on a rainy day".
And finally "crap" is spam, no?
- - -
What about a service that would let you give a special email address for Amazon confirmation emails, for example, that would receive these emails and index and categorize them. When a package doesn't arrive you go there to look for the tracking number, but only then.
Of course you can do that with tags in Gmail.
What you can't do in Gmail is to generate a single PDF of all proforma received from a given provider, for example.
Speech ---> messenger ---> letters ---> telegraph ---> telephone ---> fax ---> email ---> ???
What's good about email? You can send it at any time. It's standardized and supported everywhere. UTF-8/UTF-16. Cheap/free. Extremely fast. Extremely easy.
What's bad about email? Spam. Conversations get broken (reply instead of reply-to-all). Prioritization must be done manually, and differs between the sender and receiver. Lack of privacy (PGP is a bandaid on a severed limb). Reading and interpretation takes time. Fonts, colors, and graphics can too easily be used terribly (e.g. the "view" is broken in MVC). Micro-managing messages and contact lists. Search (http://theoatmeal.com/pl/brain/memory). Conveying context and emotions is difficult. I'm sure you can can list more.
--------------------
I've been thinking about the prioritization issue, using a sort of expensive cipher, but for humans.
In the morning, Kate in accounting gets a call from their supplier for car engines. They demand $6.7M payment on an invoice 90 days overdue, and won't make the engine delivery tomorrow unless they get the cash wired to them by 5pm tonight. Kate knows they need those engines to put in cars coming off the assembly line tomorrow, so Kate sends an urgent message to Susan, the CEO. They have the cash, all Kate needs is Susan's approval to wire it.
Susan gets the message in her inbox marked "urgent", but doesn't read it. She has 1,000 other unread messages, 200 "urgent". Susan gets through 800 messages before leaving, but never gets to Kate's.
The next day Susan finally reads Kate's message and FREAKS OUT! She calls up the car engine supplier about the late payment. "Sorry, no engines today! Next time make sure you pay us on time!" Car assembly halts, productivity bottoms, and they start burning money from idle workers and sales lost to late car deliveries.
In a small company, Kate would just walk into Susan's office and remind her. In a large company, with a CEO who's "SO BUSY", or another human who's "SO BUSY", important messages like that can be read too late. Here, Kate knew it was really important, and marked it "urgent", but because too many other people cried wolf too, Susan never got the message.
So, how can you prioritize a message, relative to all other messages someone else has queued?
Imagine an economically-consuming task. We'll use time, but money could work too. When you press "Send", a button pops up. The more times you click on it continuously, the higher priority your message will be relative to others who clicked it fewer times. Your message isn't sent until you stop clicking for 5 seconds.
So, if you click your button for 10 minutes in a row, which is a total waste of your time, your message will appear higher on someone's message-queue than someone who clicked it for 9 minutes in a row.
Furthermore, when the receiver gets the message, and see's that you're a drama queen over-reacting on something that's really low-priority, she can down-vote you. Next time, 10-minutes of button clicking will only count for 5 minutes. The sender gets notified that they cried "wolf! wolf!" and can alter their behavior accordingly.
Clicking of a button doesn't have to be the actual task. It shouldn't be able to be automated (button-clicking can), and it should be boring (playing Tetris would be disastrous, yielding every message as URGENT).
Cash works too. If you cold-send an email to Bill Gates, he's probably not going to read it. What if you risked $100 sending the message? Bill has 48 hours to read/reply to it, otherwise the $100 goes back to you. If he does read/reply in 48 hours, he can keep the $100 (don't bug me again!), or return the $100 (let's keep chatting...).
Cash is interesting though. I can imagine how it would technically work.. but would anyone actually pay? I think it depends on what sort of 'email problem' you have. I don't imagine someone's boss would be happy to get a notification that it costs $10 to send their employee an email ;) It seems this cash thing would work only for blind introductions to popular individuals.
The cash solution is interesting, but again, do you want only rich people to send you email?
I think the prioritization problem is hard because it can only be done by the recipient (not her secretary, not her colleague, not anyone but her).
But I also think that people hate to prioritize not because it takes time, but because it's incredibly boring and then stays in your mind (you remember you were bored, and you remember all the useless crap you had to go through).
But very important and very busy CEOs still watch TV or go to baseball games, so they do have time left; it's just that they'd rather watch TV than sort through their mail.
So the problem becomes: how do we make prioritizing fun?
(I have no idea how to answer that question, but rephrasing the problem is a first step).
I always thought that a simple first step would be to group messages by senders, but AFAIK the only client that does that is the concept client persona (http://personamail.info/).
The signup for the closed beta is $100 though, which is expensive just to test something.
Also, it seems that in order to use it you have to migrate to their service completely, which is a big commitment.
Having a service that progressively takes part of your emails would be less aggressive maybe.
Some thoughts:
1. Whatever the new email is needs to look and feel like email at the beginning, though it can morph into something else over time. It also needs to be completely keyboard driven and entirely API controllable on the client and mail server side. GMail's API is only so-so and a moving target.
2. It should be initially marketed at power users who can define sophisticated functions on emails or groups of emails. Those with 100s of emails a day feel the most pain right now.
3. On Stackoverflow, before you can submit a new question, it is intercepted and searched against the database to determine whether it has already been answered. Similarly, you could have something that directly links email to search of a corporate database. Before you hit send, it suggests alternate things you could look at: Google Docs, github, the general web. This search would be a combination of statistical NLP and deterministic rules set up by the system administrator, and would intercept many informational or TODO type queries.
4. You might do something much more interesting with auto-responders. E.g. someone asking for a meeting could be redirected to your calendar to self-serve schedule if they are on your "ok to meet" whitelist.
5. More interestingly, you might also allow API-hooks for trusted senders and receivers. For trusted senders within your own corporate network, you might allow them to put some kind of button or hook in the email to permit more structured responses than free text. For example, they could discharge the TODO directly without having to click a link and leave email. Or in the case of the CEO below, it would be "authorize payment". You might actually be able to make this work today within GMail with suitable iframing.
6. A properly configured mutt or Emacs gnus has many interesting features that could be ported to the browser.
Yes. We already have SIRI, why doesn't this already exist...? It must exist somewhere and we're not aware of it, no?
Mostly, I filter out 'family', 'friends', a few mailing lists, and some other random things, not to mention spam, and use "mark as read" liberally so that I always know what I have and have not looked at.
I don't really use the important markings or stars in Gmail, labels + archive are enough. Only important items stay in the inbox and they only stay there until dealt with, so I keep my inbox clean.
Is that (fun part) generally true? I'm in my thirties, and I certainly had lots of fun at my previous job (MSFT, on developer tools), but none of it compared with either my undergraduate or Ph.D. student experiences. That also seems to be true for most of my colleagues and friends --- who have done the startup sold for lots of money, big smash-hit video games, part of the i* device releases from day 1, etc.
All of the faculty I know really love it as well. Certainly, I hear bad things from the assistant (untenured) faculty, but all of the folks I meet who've made it over the hurdle seem to really love their work and have a lot of fun.
Are non top-tier universities really that bad? Or is my circle of friends and programming languages researchers just a bunch of weirdos?
I get the impression that the best American universities are perfectly fine, and then some.
Yes.
It must be amazing to have gone to a university with good teaching and electives, along side peers who actually cared.
The far side of things is the kind of college life portrayed in 'In the Basement of the Ivory Tower' [1], or 'College, Inc' [2] where it has gone far beyond just 'not having fun' and more into full fledged panic mode.
Innovation wise, I think the kind of degrees and approach (flat fee all you can learn in a semester) of Western Governors University [3] is really interesting. It seems very much like a Khan's Academy style higher education offering.
1 - http://www.theatlantic.com/magazine/archive/2008/06/in-the-b...
I do still have quite a few friends who have trouble divorcing themselves from how tough it was, some of the "unfairness", and having to do learn things they didn't (and and most cases I assume still don't) see the point in. Perhaps the same sentiments are more common/intense at universities and programs within them that are middling quality?
Over in the humanities, the faculty gave up on / let themselves get chased off the Great Books about two decades ago, what they're left with isn't much fun. The social sciences mustn't come up with the Wrong Answers, or else. And the whole notion of merit and quality is sort of suspect in those disciplines, so the joys and terrors of seeking excellence are heavily obscured.
So, yeah, those programs are a lot less exciting than they should be.
Yes, I got my butt kicked by it but every course after that was comparatively easier because of the rigor of that course and the mindset it left me with.
* Pay disparity between professor and industry; if you optimize for money, you go more industry.
* Papers are a key metric of success, not actual results (Which are hard to quantify)
* Defunding of public universities
* Rising cost of education vs. inflation or cost of living
* Students loans leaving graduates (and non-graduates) in heinous amounts of debt. It's very hard to pay your way through school now with a part-time low-wage job.
* Worsening average quality of incoming students causes higher burden on instructors and institutions
* Increasing amount of specialization; liberal arts majors can be very smart and have broad understanding, but get shafted because paying for on-the-job training is expensive.
* Training vs. education wars.
* Entitlement issues: "I paid for this course, I should get an A".
As usual, the cream is at the top, and can bring in enough money to keep things going, and can select on more than the ability to pay tuition.
Liberal arts majors can be smart and have broad understanding but so can technical people. In fact, the technical people I've met on average have been smarter than liberal arts people.
While universities have quite a few problems, universities is where most of the world-changing research has been done over the past few centuries. If you kill the university before coming up with an adequate substitute for that function, you kill the future.
That's a standard technical person bias. I'm betting that its not true on average, that a good philosopher, businessperson, or history student is equally smart, but it's not as discernible due to the falsifiable nature of technical work.
First of all - I wouldn't count being a businessperson as liberal arts.
One way you can look at it is how many people drop out of technical subjects because they find them too hard to move to humanities and liberal arts vs. the opposite. Anecdotally, for me the former number has been much larger. The latter number has been 0.
Similarly, I've had many technical friends who took liberal arts courses and generally did very well in them. I also have many liberal arts friends who took technical courses and did not do well in them. In addition, the technical people I've met have been much much more hard-working than liberal arts people. (An exception I'd make is creative arts - people I've encountered in that field have generally been extremely hard-working. Also no one I've met has moved from say math to music because they found math too hard.)
Now it might just be because liberal arts people just don't like the technical subject matter on average as much as technical people like the liberal arts subject matter. But even if that's true, one is then right to choose technical people over liberal arts people.
Lastly, I also know liberal arts majors who are smarter than many technical folk I've encountered but I'm talking about the general trend here.
People drop out from technical studies into humanities because they did not perform well academically in their technical studies. Just because a group of people can perform better at school than another group of people are not smarter. One group is better at school. And school is not an indicator as smartness. Just because a class is harder does not mean that the people who do well in it are smarter. Performance at school is not a direct indicator of smartness.
Remember, not all courses are made equal. I went to a tech school, and transferred after my first year to a high tier liberal arts school. I had a much harder time with a 050 level into liberal arts course my second year, than I did with a 400 level course on Poetry (of which I didn't have any sort of background in). I found the liberal arts courses at the tech school were aimed towards tech students - meaning, in that specific case, many of them felt somewhat watered down.
I would be willing to guess that, as a frequenter of Hacker News, you're more likely to be more active in a tech oriented circle. Growing up, and at college, I was more active among humanities oriented folk; my experience was that the tech folk were good at one thing (whatever they studied) and otherwise just played lots of video games, while the humanities students did every thing + 1.
So, all in all - my point is that anecdotal evidence is just that. Technical people can have a broad understanding, but it doesn't mean they do. Same goes for Liberal Arts.
Here is my experience in undegrad.
1. Most students simply don't care about their major's subject.
2. Most professors don't care about the subjects they teach enough to care about how students are doing.
3. Grading is aimed at punishing mistakes, rather than rewarding for learning things or doing things.
4. Guess what those things do to your group projects.
5. If you try to undertake ambitious projects or subjects for your papers, your grades go down, because of the things listed above. The most reliable way to get reliably high scores is to never experiment, never push yourself.
6. Degree requirements are pretty much arbitrary. For example, in my university, "IT" students (who weren't going to do any major programming on their jobs) were forced to take theoretical CS classes, while students in "Computer Engineering" (who were bound to do tons of programming) weren't.
7. Add to this the fact that you're expected to buy expensive useless textbooks that go from $200 to $2 once the next edition is out. (Of course, you can buy used books, but that's not the the point. The point is, they have no inherent value.)
Etc, etc. Of course there were some exceptions, couple of great profs, several enthusiastic students, some courses taught from an O'Reilly book, but such anomalies were a precious rarity.
I would disagree with this quite strongly. Most profs I've encountered love the subjects they work on - it's teaching they don't care much about.
This is mainly because teaching and research are minimally related topics which a single person is forced to perform for legacy reasons.
I do not have yummyfajitas' experience, teaching at (minimally) a top 50 US university, I've only taught high school students whose parents are paying lots for them to be there. They mostly don't care. Teaching primary school sometimes leaves me exhausted but never demoralised, middle school is occasionally like watching the capacity to care drain out of you permanently.
Anyone who does a Ph.D. for any reason other than a genuine love of research and a desire to become a professor is mad. There are fields where it's better or worse but none where a real doctorate[0] is the best way to get what you want except academia.
If you like teaching getting a good Master's degree and some highschool teaching experience before teaching at an expensive prepschool would probably be more rewarding than teaching at most state universities.
That the life of the mind and teaching the callow youth are currently bundled goods is a historical accident I devoutly hope to see ended within two decades; I'd love to see it happen sooner.
teaching at an expensive prepschool would probably be more rewarding than teaching at most state universities I've only taught high school students
talking out of your rear end? This one really alarmed me, because you're supposed to be a teacher.
Seriously, what is it with teachers who never check their facts and just banter on like they know everything? You are putting knowledge in the heads of hundreds of kids. It will be there forever. Looking at your quality of your discourse (where's that reference you were going to post?) you're not afraid of putting rotting crap in their heads, mentally crippling them forever. Knowledge (information) is the essence of your work, of your life, and of your calling. You should be nurturing it, and treating it with utmost respect. Instead, you come here, display lack of most basic education in your trade, and spew self-contradictions as authoritative statements. It shows total disregard for the trust put in you by countless hopeful parents of children left under your watch.
The worst thing? Teachers doing that sort of stuff just don't care. It's just another job, like flipping burgers. Burger flipped wrong? Don't worry, I'll make a new one. Kid fucked up for life? Twenty will come after him, I will perhaps maybe get it right some day.
Re-reading your post again and again I see more and more evil. For example:
Please expand considerably on this point if you want anyone to understand what you mean rather than giving you the benefit of the doubt.
Now I get it! You're not actually discussing with me, or replying to my post. You're grading my post! That's fucking amazing. Do I get an A? No? Well, tough luck.
You have assumed by default a position of authority over me, displaying you do not think of others as equal to you.
Let's now consider this quote: That the life of the mind and teaching the callow youth are currently bundled goods is a historical accident I devoutly hope to see ended within two decades; I'd love to see it happen sooner.
The whole post is written in simple English. Sometimes even broken English:
I do not have yummyfajitas' experience, teaching at (minimally) a top 50 US university
Sounds more akin to what someone digging up turnips for a living would say.
So you give us this cow-back ride, and then you go and put this at the end of your post:
That the life of the mind and teaching the callow youth are currently bundled goods is a historical accident I devoutly hope to see ended within two decades; I'd love to see it happen sooner.
We've got it all. Inversion, long words, archaisms, assertions of superiority, uncommon punctuation, hazy - but superficially thoughtful - statements (the life of the mind, uh, thanks, Hakunin: http://youtu.be/J376ardIXvs ). Is this the obligatory pompous verse you put at the end of every homework feedback, or am I special to receive such exquisite treatment?
Then let's drop the lexical analysis of your post and look at the logic. First off, you're replying to the wrong post. Your reply is a non-sequitur; it does not touch on points presented in my post (the single point was observational bias). You drift from topic to topic in a clutter of thoughts, unrelated to the original point. You'd normally separate those out into a post that'll be attached to something it relates to. Then, you use absolutes (Anyone who does a Ph.D. for any reason other than a genuine love of research and a desire to become a professor is mad.) that are completely inane. There are dozens of reasons to get a PhD, such as: a) helping your career b) it's one year extra, let's do it anyways c) having fun while doing it d) wanting to have a PhD like your dad and see what he's been through e) wanting to learn about specific topics only covered in the PhD course.
I seriously hope this post is much below your usual quality of submission. So.. let's spend some time together, shall we? I will now go and look through your comment and submission history.
OK, first comment: you talk about how compulsory education is... bad? Wait, are you a teacher at all? I can't possibly imagine a teacher who'll think compulsory education is bad.. unless.. yeah, unless they're as bad as you and realize they're rather doing a disservice.
Second comment: someone asks about the possible security of self-driving cars and you refer them to a novelization of Wargames. Really?
There are more comments like this.
So let's look at the submissions. An overwhelming majority (around 3/4) of submissions is under 5 points. You'd hope a teacher would have something interesting to say, but apparently this is not true in this case.
Oh, boy, how I hated teachers doing all the stuff you managed to involuntarily display in just one vile venomous thorn of a post in this thread. I sincerely hope you get fired sometime soon and can't do that to children whose mental capacity is not enough to question your bad judgment.
I rest my case.
1) Technology has made college much easier than it was 20 years ago. What used to take a few days of research in the library reading books is now a google search away. Reading has been replaced by Wikipedia. Yet, the curriculum has not scaled up in difficulty. Students spend very little time actually studying and internalizing information.
2) College needs to have a more high-school approach to teaching students, delivering an active, 40-hour week of education as opposed to 1-2 hours of lecture per day. The problem with liberal arts degrees isn't that they are liberal arts, it's that the actual activity requires reading a little bit and writing a little bit. There's no intensive, long-term internalization of the discipline.
To silicon valley, technology in education means Khan Academy. In the real world, it means a semester worth of papers written by Wikipedia in a few hours and midterms furtively answered by iPhone.
For the amount of money students are spending to attend, Universities should be able to provide a more engaging, transformative product. While this may seem abhorrent to self-starters and outliers in this community, the reality is that the demographic shift of universities has been towards exploding numbers of people with average intelligence who sincerely need full-time instruction.
If your statement is correct, you wouldn’t know it from reading the schlock that students pass off as papers even at “world class” colleges. Doing all the reading and writing clear, solid analyzes of difficult material is just as hard as it ever was. Harder, even, because there are a whole bunch of distractions that didn’t exist 30 years ago.
> College needs to have a more high-school approach to teaching students
Good god. High school is a stultifying swamp of concentrated terribleness. The whole point of college is that treating people like adults leads potentially to adult-quality thinking. Teaching college like high school completely undermines the purpose: raising a new generation of capable scholars and citizens.
> There's no intensive, long-term internalization of the discipline [in liberal arts subjects].
Just because none of the liberal arts majors you took the trouble to really engage with actually took their studies seriously doesn’t mean that serious liberal arts students don’t exist.
* * *
Overall, I think you’ve confused the difficulty of making good grades in a college program with the difficulty of doing the assigned work. If you measure your growth by how many A’s you can pull, education is always going to be intellectually unsatisfying for you.
I'm addressing the day to day experience of most students. Most students spend very little time actively learning, and the work they do is diluted by technology.
Large universities aren't treating students like adults- they aren't engaging them at all.
I'm not recommending a high school format curriculum, but the trope that students spend two hours reading/working for each hour in lecture is used as a de facto standard explaining why students are receiving a full-time education. In practice this premise is false.
Freedom and ability to explore is actually THE main advantage of the US college system.
Now slackers will remain slackers whatever "platform" you put in front of them...
I own 100s of books the content of which is in no way googlable, available via ebooks, or anything else. They are undergrad level books... nothing fancy. Your comment shows you haven't been to a demanding university.
> College needs to have a more high-school approach to teaching students, delivering an active, 40-hour week of education as opposed to 1-2 hours of lecture per day.
I thought we were talking about universities. It sounds like you want to continue being spoon-fed all your life.
> Universities should be able to provide a more engaging, transformative product. While this may seem abhorrent to self-starters and outliers in this community
Uh, how about this: we leave university degrees as only available to people mentally prepared for doing hard work on their own, and we give people who pay to be force-taught fischer-price toy degrees in bright colors? Laminated in case they slobber all over them.
But there already are universities that do what you want. Degrees from them are worthless unless the person hiring doesn't know the difference; and in that case you can bet the company is full of idiots and will fail anyways.
I mean, it's a tiresome task to even attempt to enumerate all the ways in which I feel college failed to live up to the hype they feed you before you go in. And perhaps it's because I went to a big, run-of-the-mill state school, but I typically refer to college as an extortion racket now.
I welcome any and all attempts to disrupt what I consider a corrupt and ineffective system.
-Publication pressure makes it such that research progress has become significantly harder to gauge (e.g.: it's become uncommon for results to get reproduced) across many fields (statistics, machine learning, the health sciences, which I've done work in for over 7 years at a highly respected institution)
-Balkanization of research fields where problems become increasingly exotic and narrowly focused to the point of being contrived (all the better for publishing papers of course)
-Increased difficulties in securing research funding or employment (more temp jobs though);
-Arbitrary metrics often being used to judge one's work. Example: if you do theoretical work, reviewers can reject your paper for lack of "applications"; if you do applied work, they can reject for "lack of theoretical depth" (what either of these terms mean in 90% of cases is arbitraily defined by a single reviewer who just wanted to torpedo your work).
-In some fields, entire mafia groups have formed that rigidly control the review process for that field, which means good luck to you if your work competes with theirs and you're submitting a paper/applying for grants
-Lack of accountability as to how research dollars are spent (vis-a-vis the public's interests)
-...
BTW, I say the above as someone who did a lot of enjoyable open-ended research as a grad student.
But the paper chase is utter horror. Crap papers and crap results are trotted-out as productivity in some sort of bibliometrics hell-scape.
Quantity rather than quality is part of the wrong path academia took. PG is right to focus on universities as ripe for a shake-up.
I think I'm in your camp. I left the research world after finishing a PhD, but before I did a post-doc. The first few years of grad school were great fun -- far more than any job I've ever had -- but eventually the music stops, and you need to find a chair. And that's when it starts to suck. The academic job market is dismal, and third-tier universities all over the country are staffed with graduates of top-tier PhD programs. You have to be a true exception to land a good professorship these days, and the ivory tower is littered with the corpses of the brilliant people who failed to reach the top.
I had always been eager to become an entrepreneur, but back when I was right out of college, there was no such thing as YC, and definitely no such thing as a college kid getting venture capital to try out an idea. So I went to grad school. Years later, I'm looking down the barrel of a crappy academic job market, but the startup world had become much more vibrant. So it was compelling.
That said, there's absolutely no way that your average startup is going to be as intellectually fulfilling as graduate studies in a field you love. Most startups are doing boring CRUD apps that take little technical ability or knowledge to produce, but a good grad program in computer science will let you work on the technology that startups will be hyping in 2032. There's no comparison.
I think that's a myth. The fields I studied aren't any narrower today than they were two decades ago -- the fields I studied didn't exist two decades ago. They barely existed ten years ago. And now there's a vibrant field of research. The same is true all over the biological sciences, computer science, chemistry, biochemistry, bioengineering, etc. The fields where a PhD means something are constantly re-inventing themselves, because the frontier of technology keeps getting bigger.
In the 1980s, Rodney Brooks at MIT was doing some interesting academic work with simple, toy robots, while everyone else was building thinking machines and planning systems and industrial robots. Twenty years later, and there are multiple companies commercializing the technologies emerged from that work. Some people might look at iRobot and say "what an innovative startup!", but I look at them and see a company that has done a good job of commercializing 20-year-old technology. They've definitely refined it since then, but the core technology was invented at MIT, in the 80s. The same applies to nearly every "hot" technology that startups are pimping today -- dig a bit below the surface, and you find an iceberg of academic research.
And that's what I mean when I say that the startups of 2032 will be hyping academic research done in 2012. It's all a matter of perspective.
"Only way to get a product visionary as the CEO of a company is to start it and not get fired."
From the same part in the article,
"None of the existing players will – not run by product visionaries."
Later elaborated on as: "just have to be better than samsung, hp, motorola – not so hard"
Doesn't this in itself bring an interesting question? Why is that those companies cannot innovate or have visionaries? Surely, they must. Maybe that's another opportunity though. How to find the visionaries in your company and reward them for doing so?
The best a large company can hope for is a skunkworks.
I'll have to dig up my BeOS Bible this weekend, but I though the BeOS solved this problem over 15 years ago. I have been floored for years that chip makers keep telling developers to write code to work with multiple CPUs. If those developers are OS developers, then fine. But application developers? No way. This is the job of the OS!
Separate the UI thread from main() (the thread doing the core app logic).
Yes, many OS's suggest spinning up a worker thread, but by default an app is single-threaded. The default in BeOS is 2 threads.
To the original poster, I'd like to point out that BeOS didn't (and Haiku doesn't) use a different compiler. What pg was saying was that we need a better compiler + language. Right now the programmer must manage every detail of the threads. A few abstractions are available - OpenMP, Apple's Grand Central Dispatch, etc. They don't really make highly parallel implementations though. The field is ripe for disruption.
Look at Verilog/VHDL for an example of a fully parallelized language.
It's an interesting idea. It would be trivial to write boilerplate like this so that when you create a new .c file, it will start with a basic multi-threaded application.
I remember seeing some parallel FORTRAN code in the late 80's with things like loops explicitly marked for parallel execution and other tricks. The downside is that debugging mutable state across multiple threads is a bitch. It's not something I would wish to my worst enemy.
> Look at Verilog/VHDL for an example of a fully parallelized language.
I have been playing with the idea of using Verilog or VHDL to express computer programs... Did nothing, but it would be an interesting thing to test.
1. dinosaur egg – make a search engine that all the hackers use. (top 10,000). - Like Google Code Search? That didn't turn out too well.
2. inbox is a todo list. email is the protocol for putting stuff on it. - What? I use email to converse with people, not to list what I need to do. Occasionally I will send myself a reminder mail with a list of TODOs, but my email inbox is definitely not a todo list.
5. his friend from apple: there will be no new good stuff post-steve jobs. - Is his friend Tim Cook? Isn't Apple famous for secrecy even inside Apple or do all 9000+ employees in Cupertino know what's in the pipeline?
Etc.
PG wishes that aspect be more explicit in order to gain more control over it.
It's funny that I read this now, during my lunch hour I was pondering over how email is broken and needs to be replaced I actually came up a few few innovative ideas but there's more to it. I don't however think that our curent Email system will be around forever, instead of a static thing that you receive it will will be more like a feed that can be paused and saved off to different tasks that could theoretically automatically handle the task for you. (FEED: Credit Card Statement just posted - You: Siri pay the bill from account XYZ)
Duckduckgo is sort of trying this approach. It isn't taking over the world, but I think it's having reasonable success.
> Occasionally I will send myself a reminder mail with a list of TODOs, but my email inbox is definitely not a todo list.
I think many people (myself included) feel their inbox is a list of things they need to respond to... which is somewhere between a conversation and a to-do list. But a full inbox can feel like a list of obligations.
> Isn't Apple famous for secrecy even inside Apple or do all 9000+ employees in Cupertino know what's in the pipeline?
His friend may be talking about the culture of Apple rather than the list of current products. I'm still skeptical of his friend's claim, but that seems a possible interpretation.
Google Code Search is not the only example of a hacker search engine. StackOverflow is a better example and is turning out fantastically.
He's not saying it's a todo list for you to send in your own todos. He's saying that a lot of the mail you receive includes things you need to do.
It's not that his friend has seen Apple's pipeline. It's that his friend does not believe that Apple will be able to keep creating breakthrough products like iPod, iPhone & iPad without a product visionary CEO.
It was neat that such a thing existed, but every time I used it, I really had to get creative about getting it to cough up useful results for me. It certainly wasn't in the same league as Google.com search, or Gmail or Google Maps.
Whenever I tried using it, I could think of a million ways to make it better, easier, more effective, and so on, but they never seemed to do anything with it. I don't know how you'd monetize it, but I think it wouldn't be too hard to make something far more usable than GCS.
don’t try to identify a precise thing in the future. better model: columbus. “there’s something west. i’ll sail westward”.
I bristle at seeing Colombus propped up as a model. The guy was wrong (he thought the earth was smaller than it actually was, so attempted to sail west as a shorter route to the East Indies), wrong (he discovered the Carribeans, not mainland America), wrong (he still thought he found the East Indies, i.e. south-east asia!).
But alas, actual truth carries less rhetorical power than popular myths.
But I understand that PG's real point is "trust your instincts, don't listen to the naysayers, but adapt to the circumstances".
I don't think my job is anywhere as risky.
Paul Graham is now a pundit and member of wall st west (Silicon Valley). He has a vested interest in pimping his product - Young impressionable developers. You can take anything he says with an iceberg sized chunk of salt.
It's clear what Paul's values are - he's been writing about them since long before YC. I wish someone like him had been saying such things back when I was a "young, impressionable developer".
In fairness, as he stated himself, he wanted to write a proper essay first about it so was attempting to come up with something on the fly.
I felt the proper response should have been something more along the lines of: Copyright only exists because society in general recognizes that right. So if society in general no longer believes it is a right, then business models will have to change to reflect that.
I personally don't know what the right long-term answer is, but clearly, since you can copy bits for "zero" cost, it's going to be even harder to convince people to pay in the future as they come to that realisation collectively.
http://www.qualcommtricorderxprize.org/
And the idea of the X-Prize seems to be mostly in-line with the spirit of Y Combinator. I only last night saw all the current X-Prizes and am pretty impressed.
I like this, or it at least validates some of my thinking. All of my startup ideas are small and modest, something that will help me and maybe some people like me. But I have no idea where the idea will take me...
So I guess a question is how to distinguish between marginal niche, and something small in a huge market. I guess it's market size and gut feel of potential.
I think what he actually meant is that if you logon to your mailbox, you see threads of stuff to do. Hence, your mailbox with names, subjects and content is really a "todo" for the day/week/etc.
I think, in the near future, we humans will be a creating protocols on our own. We will be creating our own protocols and "pushing" data, instead of "pulling" it from the web, like it is right now.
In my spare time, I am actually working on a paper covering that. I call it "hapi": Humans with Application Programming Interface. Will be glad to share more when its done.
I'm not an expert in this field by any means, but I think this idea is not so "frightening".
There was rumors in 2006 that AMD tried to develop "Reverse Hyperthreading". No official notes though.
Edit: Intel Anaphase: http://newsroom.intel.com/docs/DOC-1111
- POP3 is completely braindead, so moving on...
- Imap doesn't natively expose many gmail features. imap doesn't know about "Archiving" and "Labelling", so you have to map those to things like deleting and folders. There's an impedance mismatch.
- Exposing a search interface with all the functionality and responsiveness of the web app is painful. Sparrow does it nicely, in my opinion.
- You sometimes can't login and don't know why. Sparrow does this badly. Once, Sparrow just started putting red dots on my account and all it said was "authentication failure". I actually had to dig in to Sparrow's (thankfully provided) JSON logs to see that the error message was that "non web clients are locked out on your account". In turn, I had to just KNOW that this meant I had to do Google's unlock captcha. But they don't make that easy to find, I had to google it. Imagine my mom trying to troubleshoot that. I don't think so.
- Enabling 3rd party access to gmail via Imap is not a trivial exercise for the non-technically inclined. Even more so if you have two-factor authentication turned on. Imagine my mom trying to set that up. I don't think so.
- Imap and SMTP aren't the most trivial of protocols to implement correctly and completely.
- Stupid little things. Like, right now if I want to have my Google Voice number ring to Google Chat, I have to have my GMail window open in the browser (someone correct me if there's a better way). If I already have it up just so I can receive phone calls, why do I ALSO want to have a standalone email client open?
>if so, prove it if not, the expected value of working on it might be really high.
parallel complexity theory: https://larc.unt.edu/ian/books/free/pct.pdf
In a society where engineers are arguably playing an increasingly important role, I think this outlook is too limited. This perspective is what I would expect in a CPA convention. If engineers are so smart, powerful and essential, is this what you want to inspire them to do? Have them make cooler gadgets, easier "content consumption experience", better to-do lists to make people more "productive" and, lastly, possibly provide better health so that they would live longer to buy more gadgets and "consume" more "content"?
If engineers are so smart, powerful and essential, why not inspire them to find ways to alleviate poverty and promote justice, education and equality? Well, you say, this is not a job for engineers. Well, if engineers are expected to think of new ways to book sleeping arrangements or sell clothes, their insight can surely be turned to loftier goals. I don't want to be caught saying anything positive about Facebook, but for better or worse, they have changed the way people around the world communicate with each other, compete with each other, and see themselves as part of society. So, yes, engineers are certainly capable of achieving some great feats.
Many lawyers make lots of money. Lots. And probably many of them became lawyers for that purpose. But It's hard for me to imagine - and maybe I'm wrong about this - but it's hard for me to imagine a well known lawyer delivering a keynote speech at a lawyers convention, where his main points are tips for making lots of money. Lawyers may be hypocritical (and they are), but at least many of them separate their business goals from their sources of inspiration. So why should engineers be inspired only by business tips? Is money the main source of inspiration for us? Doctors and lawyers make money, too, but they aspire for - or, at least, are inspired by - more. Even writers, directors and actors aspire for more than making money, or even "delivering content" or providing entertainment. They dream that, perhaps, some of their work might carry some Truth. Give people a new perspective about life, maybe even change people's lives - but not by making them simply more "productive". Peh, productivity - if there ever was a more overrated and servile word.
Anyway, we should aspire to do the same. Fellow engineers, do whatever makes you feel good about yourselves. Do whatever you need to make lots and lots and lots - piles - of money. By all means, solve people's annoying problems; that's a great advice. But don't aspire to that. You should dream about solving humanity's serious problems. You should dream of making this a truly better world - in every respect.
I found that one of the worst and hardest mental habits to kick was the notion that I had to do "big, important things". Because big, important things are ill-defined, and your success metrics aren't clear. It's very easy to kid yourself into thinking that you're promoting justice for all, right up until you're 20 years in and the world is still not a just place. And then you have a mid-life crisis, think you're a failure, and give up.
It's much better if you can point to specific and yet often underwhelming-sounding instances of "I made this person's life better" or "I let this person do something he otherwise would not have been able to". A journey of a thousand miles starts with a single step, and all that.
Anyway, as I said, I don't really have a problem with anything he said in particular. Some of the things actually made me think. My only issue is the premise that engineers should be inspired by business. Case in point: "it would be great if a startup could make a lot of cpus look to the developer like 1 cpu... the expected value of working on it might be really high. why high: web services. programmers like convenience." I would have said that with a compiler like that, scientists would be able to design new drugs more quickly, or that a manned mission to mars could be done more easily, or... I don't know. But his example (at least as it appears in the notes) is web services.
Again, he was talking about how to come up with ideas for a start-up, which is a business, I guess, so the talk was very much to the point. But even if Mr. Graham believes that start-ups are where smart engineers should spend their best years and bright minds because start-ups are what makes the world better - and he does indeed seem to believe that - I think he should have at least discussed why it is that bright young people should strive to build high-risk fast-growing businesses, rather than, I don't know, go work for the government or an NGO. It is as if many engineers just assume that the best path for an accomplished engineers is in the fast-money business. Engineers should be agents of freedom, of equality, of innovation and even social justice. I just don't like it that nowadays engineers in certain fields are automatically associated first with money, as if they were stock brokers or something. This is important because history has shown that smart as they are, scientists and engineers were often very easily manipulated into serving some terrible causes. They are easily manipulated because they are drawn to solving problems rather than thinking about whether they should be solved in the first place.
For example, an easier delivery of entertainment "content" is a problem for an engineer. But is it a problem worth solving? From a business perspective - certainly. From a human perspective - possibly. I don't know. But young engineers eager to learn from an experienced colleague are already trained to think about how to solve problems. They're even trained to find problems to solve. But they should be inspired to think about why it is that a problem requires solving, and to consider whether there are other, more worthy, undertakings. Again - in the end they'll do whatever it is that they're doing now. Only that shouldn't be their inspiration.
Also, big things don't have to be well defined. They should just hang up there as a source of inspiration. Big things, not rich people.
You (correctly) point out the places where capitalism comes up empty, if capitalism is our implicit system of determining the worth of goods and services. Apart from the environment (which seems unaddressed by the major economic models of the world) or the other causes you mentioned (unaddressed by capitalism, at least), I suppose it's fine to rely on the adequacy of money as some measure of the wealth we add to the world (http://www.paulgraham.com/wealth.html).
Even if the world is being eaten by software (http://online.wsj.com/article/SB1000142405311190348090457651...), the transition to using software happens fastest through startups, I'm coming to believe. I used to work in a life-sci research lab, and there's just an exponentially growing divide between data growth and the tech resources to handle it. NGOs and government can be disorganized and unmotivated, so working for them is not necessarily going to be the best use of your engineering talents. But that might be aside your point, i.e., try to work towards noble goals.
Lastly, and most importantly, I find the model of the Aravind Eye Hospital as truly inspiring and business-model bending: a company set up to help the ones who are the _most_ in need. The ones who can least afford it. And yet in the process, generate income that goes back to serving and accelerating the mission. _This_ is what "social enterprises" ought to be: http://infinitevisionaries.com
The what will be cooler gadgets and "increased productivity" or "streamlined processes" only if the why is money. Money is perfectly OK as motivation; it's just so... pedestrian.
"To build a house, put a brick down on the ground and see where that takes you. What's a blueprint again?"
These are very integrated problems spanning a wide range of disciplines to solve. Pure engineering geekery isn't going to solve them, and engineers need to recognize that we limit ourselves by going deep.
Anyway, at my day job, part of our company's work in the past decade has been to get electricity to villages that never had electricity before. That's pretty awesome. :-)
I'd be curious to see the percentage of deaths caused by delayed diagnostic.
Genomics prices are heading down, let's make it easy, fast, and integrative.
Umm, search is just fine thank you. I could not survive without Google currently. The only thing I don't like about them is the lack of good customer support and the way they tend to develop unintuitive things with poor documentation (appengine was hard as hell to develop to, Android is getting better, but still they don't have a Rails-style "get up to speed" doc- they have a long way to go) But search is not their problem or mine.
> 2. replace email
> inbox is a todo list.
No, it isn't. It's a method of asynchronous communication and file transfer. If you're using it as a TODO list, you're doing it wrong.
> powerful people are in pain because of email. that’s an opportunity.
B.S. Email has survived Facebook, Twitter, Google Wave (cough)...
> whatever you build, make it fast. gmail has become painfully slow.
Maybe on a Pentium III. It isn't slow for me though.
> 3. replace universities
Good idea, but won't happen. The problem is any business/institution that gets continuous revenue without having to be accountable. Universities suck because they don't have to provide what they are needed for, they only have to compete with each other. The service the universities should provide is the preparation of its students for the betterment of the world. However, what this means is debatable, and the hippies of the 60s that grew up in an environment where they didn't have to work their asses off (like the teenagers of the depression era) are the ones teaching our kids how to feel better about themselves by building a hut in a 3rd world country.
> 4. kill hollywood
The studios are the only ones that really know how to produce. Music, T.V. and movies are all about production. The YouTube era won't last forever. It is only a matter of time before they get a full handle on things again.
> 5. a new apple
Apple is still on top of the consumer market. Solve problems that need solving in the way you feel is best. Don't work hard to be something else. Apple didn't.
> 6. bring back the old moore’s law
>...it would be great if a startup could make a lot of cpus look to the developer like 1 cpu.
As an analogy, why don't we make Ruby look like 8086 programming? Parallel computing is different- you can't solve problems with the same mentality.
That means most business users of e-mail are doing it wrong. Take a sample of heavy e-mail users (managers, sales people, marketing, etc.) and ten out of ten will tell you that they're inundated by e-mail, and that it is a defacto todo list. E-mail is a giant problem for everyone who needs to communicate with a large number of people professionally, and "you're doing it wrong" is not the answer. Using e-mail as a todo list is the path of least resistance, so it's the only viable option for most people right now, and it sucks. It needs to (and will) get fixed.
In general, it's easy to go through any list of ideas and dismiss them. All of these ideas seem sensible to me. They're vague by design, many of them will likely take a very different shape, but it's a good list of general directions to explore.
TL;DR don't be a cynic, be a builder.
Beautifully put.
I don't think pg is asking for help with his email - for MANY people it IS a todo list. He's outlining opportunities for future startups - email IS an opportunity precisely because people are "doing it wrong" - you snarking like you disagree with him but it seems like you actually agree. He's saying there needs to be a better todo list system for EVERYONE email can still fill the role of asycn comms and file transfer.
>>gmail is slow >it isnt slow for me
Maybe I'm wrong but I assumed he was meaning this to be it takes a lot of time not it has poor performance - again seems like your are just trying to negative.
>> 5. a new apple
>Apple is still on top of the consumer market. Solve problems that need solving in the way you feel is best. Don't work hard to be something else. Apple didn't.
He's talking long term - he's talking about apple when they were in a garage - he's not saying they are dead - he's saying with Jobs no longer at the helm there is less likely to be next generation vision - so again there is an OPPORTUNITY - someone will be the next visionary - maybe it could be you.
>> 6. bring back the old moore’s law >>...it would be great if a startup could make a lot of cpus look to the developer like 1 cpu.
>As an analogy, why don't we make Ruby look like 8086 programming? Parallel computing is different- you can't solve problems with the same mentality.
I don't understand your logic here. PG is talking about abstracting the problem away from the developer - you think the future of CS is that we will no longer make things abstract?! I think PG mentions it because it would be HUGE if this could be done - it certainly is difficult - he's not advocating any kind of mentality - he's just saying there's an opportunity here.
I interpreted him as implying that writing code for parallel processing should happen in such a way that we don't have to think of it. But that is the whole problem. Developers are ignoring the complexity and opportunity that exists in understanding parallel computing. Development languages should evolve to embrace the opportunity that exists within understanding what is going on, not hide it. We didn't keep programming in BASIC since we were kids as our primary language, nor did we continue to use assembly. Things evolve, and, as developers, our thinking needs to evolve to support multiple CPUs and parallel computing; we don't need someone's help to wash over the fact that things have gotten more complex. We need to understand that complexity before attempting to abstract it.