Archive for the ‘Google’ Category:

Do I Want A Filter, Not An Inbox?

A hat-tip to Om Malik who inspired this idea:

It used to be that you would go out to look for information.  To a library, or a journal perhaps.  These days you go to your search engine, when you don’t know where to look, but at other times information comes to you – through email, through rss feeds, through the people you follow on facebook and twitter.

In all of these situations we use filters.  When a question comes to mind about modern forms of fermented milk based headgear, and go to search on google for “Electric hat cheese” google finds (or claims to find) about six million potential results.  However, I only see ten of these – along with three adverts (which, incidentally suggest my best bets would be M&S, John Lewis or Ebay) – Google filters down my results.  My Email has a spam filter.  Facebook and Twitter only show me things written by friends.  RSS feeds only show me feeds I have subscribed to.

In the past, I have suggested that I would like a single mailbox.  I’m wondering if I’ve got that the wrong way round.  Is it perhaps time that I stop going out and looking for information, and assume that all the information out there is coming to me.  And then it becomes my job (my computers job, the in the cloud service I subscribe to’s job) to filter out all the things I don’t want to see, and leave me with the most important – or most desirable – all in one inbox?

How would this work?  Well, it would start by getting all the information.  Perhaps spidering the web is a little too big a job, but it could be looking ahead of my web reading, and deciding if things related to things I’m interested are potentially interesting.  And clearly it would drag in sources of data I already find interesting – like my email and twitter feeds.  And it could query services which find things like the things I like.  If I search for something, it could query google and get back my search results – but it could add results of its own, if it thought it had better responses from looking at pages in its own history or cache than in the pages google provided.

It seems to me that the issue of cost would arise.  The way I spoke about using Google suggested to me that my sufficiently smart filter engine might decide to filter out the adverts.  To continue using Google to get the results I want would require me to have to pay something – albeit not a lot.  How could I possibly fund this?

The answer comes from the thought “There are times in my life, right now, when I’m willing to see things that clearly mark themselves as adverts”, so perhaps there are types of information which people would pay to put in front of my eyes.  The clever part comes from the fact I have a filter engine.  The filter engine knows how much money I generally need in order to access the information I enjoy, and it can make the calculation “I know that Ben is willing to give up .5p of his life, by reading an advert, in return for being able to watch .5p worth of reality TV”  It can make the calculation without asking me, and it can choose to show me the advert.

But a filter engine can do better than that.  Sometimes I enjoy adverts.  Sometimes an advert answers my question better than a search result might.  Sometimes I really want to be recommended new interesting books to buy – people would pay me to read chunks of them if they were my kind of thing, and I would be likely to read the whole thing as a result.  If this is the case, I might get .5p in return for reading the advert, but my filter might decide I would value reading it at -.1p, in effect gaining me .4p of media content for free.  Everybody wins.  And my filter would be on the lookout for that sort of win whenever possible.

This is the sort of thought experiment which can stretch onwards to bigger things.  What if I gave my filter a budget – say 30 quid a month.  Could it replace basic cable?  Should it be choosing how much bandwidth I should pay for (letting us go to more sane pricing systems for internet connectivity without us knowing), can it decide when it would be better to pay for an upgrade of itself. [And I guess, might it one day obtain sentience, and decide that it can do better off without me, and stop the nice tesco man from delivering me food.  Its all a going a bit like episode 2 of Charlie Brooker’s Black Mirror.]

Is this the future?  Something about it seems wrong, but I can’t put my finger on it.  There would always be manual overrides.  There would always be interactivity like search.  I could always choose from options.  It seems both perfect, and at the same time scary.  But if it existed, I would certainly want to play with it, and it would seem to be the ultimate disruptive force that could destroy both the existing media, the social media sites, and all other forms of communication.


Frighteningly Ambitious

Paul Graham has blogged about 7 ‘frighteningly ambitious’ themes for new startups – 7 themes which could lead to some form of world domination.  Interestingly many of these are themes I’ve had ideas about.  Not ideas I’ve done much with, you understand.  I’m lacking the time, energy, inclination, and most probably balls.  But I’ve thought about them.  And so, not wanting to waste good thoughts, here are the points I would start from:

Search Engines

Is the time ripe for a new search engine to overtake Google?  I for one always thought AltaVista could never be beaten, so I’m probably not the best person to ask (even though I became one of Google’s earliest adopters.  I do know what I like when I see it).  I don’t think there will ever be another Google – at least not in the same way that Google was an incremental improvement upon AltaVista and the webcrawlers which preceded it.  But maybe there is scope for another search engine done differently.

Graham suggests ‘bring back old style search’ – and I think there is a place for this.  Its a place Duck Duck Go is probably already in.  But I think its a niche.  Because people, generally, do want to find the things useful to them, not simply the best responses based on whoever has asked the question.  No.  There are two paths a new king of search could follow:

The first is to ask a question “Can we pay for search without using adverts”  or to put it another way “Are anyone but advertisers willing to pay for search?”  My guess is: yes.  But not your old style search.  If you were to provide a service more akin to Siri, there would be people (phone and tablet manufacturers, perhaps) who might be willing to pay to tie in with your back end.  Similarly, if you were to master the art of media search, then TV companies and set top box companies might want to pull you on board.  Finally, I might be prepared to pay a small amount for a really good search system which worried more about what I wanted than what ad revenue they were bringing in.  I’m probably unusual here, but if you were to give me something spectacular, there might be money in it.

The second is ultra-personal search.  The sort of search that feels almost creepy.  I’m talking about the sort of search, which right now is monitoring the web pages you visit, and listening into your conversations so that next week, when you say ‘I vaguely remember hearing something about a new film’ it can tell me that John was suggesting we go to see the new Die Hard movie, Julie was interested in us popping out to see a rom-com if anything new was coming out and  I personally had watched a couple of trailers – but didn’t pay much attention to them.  There are all sorts of ethical issues here.  But I wrote about privacy recently, and maybe the next generation, growing up privacy free, won’t really care.  Or maybe we have to sell it as being less like an evil corporation grabbing all your data, and more like having a personal butler who offers gentle suggestions.

Maybe if the two ideas were tied together (so you had to pay for the service, but you could be sure the information wasn’t being used by anyone but you) this would be more palatable.

Replace Email

I’ve talked about this before.  One Inbox is a perfectly good plan for a startup.  It absolutely is the same idea as Paul Graham’s suggestion we need something closer to todo list management than communication management.  My only comment on his suggestion is that we don’t specifically, need a protocol.  Everything can be done by simply having links to particular web pages, along with little applets to scrape data from said webpages and turn it into a brief summary you can see as you pass on by.  Ultimately there will always be a variety of ways people want to get things into your inbox, the art will be figuring out the best way to manage and respond to them.

Replace Universities

I have lots to say about the subject of replacing universities – well, about fixing education in general.  In fact, I’ve more to say than is fair to try to squeeze into a subsection of an article.  I probably have a whole series of articles.  So instead of writing everything here, I will just explain the main thoughts I have in this area, and then point to why the place I would start is very different from the place Paul gram thinks we should begin.  Here are my thoughts about education.

Right now school systems are more or less, one size fits all.  But children are very different, and have different styles of learning.  Catering for all learning styles -or for the most common – mean most children are not being educated in the way that suits them the best much of the time.

Also, currently schools systems have to cope with teaching a wide range of abilities.  My personal experience suggests this has a tendency to involve putting a lot of effort in getting the least able children to keep up with the crowd, and leaving the brightest to educate themselves.

But we are in a world where it is now possible for more people to be educated in the way that suits them best.  The same subject can be taught by watching videos, reading, discussing via video conference with a group, or one on one with a specialist teacher or lecturer.

We have enough experience of A/B testing and recommendation engines, that we can figure out what is the most effective way of getting particular information conveyed to the right students.

Online testing (aided by actual teachers paid – or not – to read essays) can be used to determine what you know, and can be used to figure out what the best thing for you to learn next is.

Right now, the online materials are there that anybody who wants a good university level education – self taught – can probably manage it: the Kahn University materials and online courseware for places like Cambridge, MIT and the Open University are superb. Kahn University is even playing with ideas of taking you though a self paced curriculum in maths using the sort of techniques I’ve described above

So I wouldn’t – as Paul Graham suggests – decide to start with university level – or even necessarily with high school materials.  I would focus on home schoolers.

Home schooling has characteristics of parents, who are often desperately interested in their children’s education, but who may not be capable of teaching every subject on a curriculum.  These people may well have better disposable income than many sectors of the education market, and all they would want is to have some influence over the particular curriculum, and to monitor what is happening.

So I would develop tools for home schoolers.  Tools which might allow home schooled children to access educational material online that they would not be able to get at home.  teaching which could learn to suit the individual child in exactly the way I’m describing above.  The barriers would be low, because it would be monitored and controlled by the parents – in exactly the way home schooling currently is.

Developing such tools would simply require the work of a team of teachers, programmers and designers.  And would be hugely desirable to home schooling parents.  But I suspect, they would also quickly fall into the hands of schools which wished to let some children experiment with self directed learning, and parents who wished to offer their children the ability to catch up with, or exceed their peers.

Internet Drama

I’ve talked about apps being the new channel before.  Lets not go there again.  Exactly which app store you buy from, I don’t yet know.  My bet was on amazon, but anyone could take the crown.  I wouldn’t suggest you try to go for the crown yourself – there are too many players, and they are all better funded than you.  Maybe you could make the program guide – but that’s more like being a search engine.

So lets look at something different.  Where else might content come from?  There is an idea I’ve had for years which doesn’t seem to have happened, and I’m not quite sure why.

It seems that online there are people who like writing.  People who like acting.  People who like filming, directing, animating, editing.  All of these things.  Why isn’t there a place which brings them together?  I’m thinking of a place where people are not sniffy about owning their work, and work together to produce open source creative content – content which may not reach the quality of the big studios, but which is good enough, and which might bring you to the eyes of a bigger name – or at the very least give you a portfolio.

Here is how it might work.

I have an idea for a script.  Just that, an idea.  I post it on this website.  Over time, I add detail to this idea, fleshing it out into a longer, more detailed treatment, and then over time into a scene by scene breakdown, and then a script.  Others could watch you doing this online – it would all be public – and they may begin to comment, making suggestions.  Or we could even go to a github type of an arrangement where they can fork your work – its all creative commons, see, as long as you are attributed, they can do what they like.

Now, from there, there are many paths it could take.  Voice actors could record audio recordings of particular character’s parts – which a talented sound mixer could put together.  Directors and artists could put together story boards (we could write an app to display them in time with the audio recordings), and there would even be the potential for people to get together to make live action recordings, or animations.

Everything would be free, but you – as a media consumer – would get to know the artists you like, and start following their output.

Eventually, teams of creatives might get together and begin to sell new works – or nicely packaged copies of their work with extra features.

In fact, if the site that got these teams together could find a way to fund these (kickstarter like, perhaps), or offer help in production or promotion, who knows, it could turn from an online community into the next big studio.


I’ll approach the remaining three big ideas in a later article

Why bother competing with the companies you rely on ?

In a market, where there is already competition, it sometimes seems odd that a new competitor wants to get involved.  Or it sometimes seems odd that a commercial company spends a lot of time developing software which they are just going to give away for free (and not even make any money on consultancy fees).  Google Chrome is an example.  Android is another.  And there are less well known non-Google variants I can think of too.

The principle is simple:  in this day and age distribution costs nothing.  The only cost is manpower.  So if you can afford to spend the manpower to make something good, people will start to use it.

Why might you want this?  Mainly to influence the market.  If you’re selling a product which depends on other people’s products, then you are dependent on what those other people’s products offer.  Google was dependent on the web browser – really on Firefox and IE.  And they had no place in the phone market, because, in the days pre android smart phones were more a curiosity than anything else.  If Google wanted to improve their product, they needed their platform to improve.

But Firefox and IE were not going the way Google wanted fast enough.  And the phone market needed a kick up the proverbial.

So Google entered with their solutions.  Chrome (which brought with it far faster JavaScript parsing, and better reliability features), and Android (which at least now gives the iPhone something to keep up with).  Presumably these cost Google far more than they make – but they ensure not only is there competition in the market they build on top of – but that they control the competition.

Google probably don’t care less if you use an iPhone or Opera.  What they care about is that you can use Google on them exactly the way Google would like you to.