Friday, February 06, 2004
Replication Wars, Pt 343
In Australia, Sharman, Brilliant, various ISPs and the homes of their execs
get raided by record company 'investigators'
... ostensibly looking for copyright violations. The government gave the record industry the go-ahead.
Absolutely mind-boggling. As someone whose origins are from the Third World, and who's spent a lot of time there, I can tell you - this is a road that we do not want to travel down. Unbelievable.
Ummm...from the link below:
"...Therefore, IEI is busily transforming all of its advanced neural network paradigms into TCP/IP based systems. The overall intent is to convert many, if not all, of the TCP/IP nodes on the Internet into functioning neurons.
The resulting freethinking entity will be capable of introspecting upon all human-originated content residing on the Internet and World Wide Web, and from that knowledge store creating new ideas and strategies that will inevitably transform our thinking and our planet".
Does this sound crazy? It's not (completely) - it's pretty much within the bounds of current technology. Whether a 'freethinking entity' would result is anybody's guess.
It probably sounds a lot scarier if you don't know much about neural networks - the computational 'neurons' talked about above are hugely, massively, primitive compared to biological neurons.
Read the whole thing here.
Nice piece about perturbing neural networks with noise to recombine concepts and come up with new ideas. Here are some specifics about the neural networks and an example - I recommend you check out the latest patent and the original patent as well.
Bill Gates is now pushing for pay-for-email. Is this sensible? A quick review of my take:
I think spam is only stoppable by economics, not technology. That's because technology is brittle, and easy to hack - but structures built on deep understandings of the microeconomics underlying spam can provide massive disincentives to spam in the first place
In the fall, I posted a detailed plan based on spamonomics - I don't have time to find the link right now. The crux was that you would put a (small) deposit in escrow, which would only be drawn against if collaborative filters or other simple technologies discovered you were spamming.
Even then, the curve of the penalty function could be altered that spamming your 100 closest friends would only cost you 10 cents on a deposit of 10 bucks - but spamming 1,000,000 would cost you exponentially more.
The thing I like about this system is that good behaviour isn't
penalized - there's no disincentive to send regular old emails - like there is in billg's scheme. If you don't misbehave, you can withdraw your deposit at any time.
Of course, then there's MS's strategic angle - you can bet that if MS is the one controlling the point of contact (the billing engine), email as we know it will come to a grinding halt - because the price will (relatively) go through the roof.
Thursday, February 05, 2004
Toys like the Swipe toolkit will fundamentally change the economics of the privacy market - by destroying the monopoly information brokers currently have, and creating markets where there are none. Stay tuned to to things like this.
Wednesday, February 04, 2004
nthoctave has some fun posts - check it out.
Are centralized, syndicated RSS ads the next big model for the Net? No, because RSS is going to fail, partly because as the value of the standard has increased, it's beginning to fragment.
And also because the economics of this are backwards - almost anything you can read in RSS you do partly because you don't have to pay the costs of seeing HTML noise - like ads.
It's amazing that it's taken about 5 years for the UK mobile industry to resort to such basic tactics as trial periods.
Why? The Reg cites 'product diffusion curves' - but I don't think these apply to tech products anymore. The answer, I suspect has more to do with path dependence - the MNO's in the UK locked themselves into strategies which didn't line up with tactics like this - and tipping points - this is probably one.
It's reports like this - statistical noise dressed up in the guise of meaningful quant data - that have led to the failure of branding.
Read a bit and you'll see my point: this report is basically telling us that brand awareness equals short-term strategic success. And the causality goes the wrong way for the unfortunate authors.
Here's a nice presentation from Apax about investing in energy markets (and barriers to growth for the PE/VC industry in Europe).
Monday, February 02, 2004
Someone's a little upset that iTunes is the same old b-model in a nice new outfit.
Score: MyDoom 1, SCO zero.
Please note the correlation between virus damage, market emergence and bounties offered - just like I predicted. No, it's not proof, but it doesn't take a genius to understand why bounties help form a market for viruses. See my article (right) for more.
This article in the NYT thinks that European and Japanese MNO's are coming to the States because there's no growth left in their own markets. That's (obviously) far from true.
I think the real reason is that American MNO's have done a nice job of opening up the mobile market in the past couple of years - they actually learned from the mistakes that their European counterparts made (and continue to make). Like obtuse and opaque pricing, idiotic marketing, feature creep, and, most importantly, price-gouging.
The reason there's no growth in mobile markets in Europe (and, to an extent Japan) is because MNO's have got the 3G proposition massively wrong - pricing, marketing, everything. I recommend you look at FOMA diary if you want to know more.
If there's any industry that's going to grow discontinuously in the next decade, it's mobile communications - it just may not be the now-incumbent mobile MNO's that see these profits, because they have absolutely no strategic insight.
the NYT gets the coming search wars all wrong - open-source search is a much more potent threat to Google than MS. For all the usual reasons - but the most interesting one is that firms are actually learning to harness user innovation.
Google has been especially good at it (until recently). This, more than anything else is the real source of their advantage - not some kind of vague (and dubious) 'technological superiority'. So: can open-source search really exploit massive user innovation, to out-google Google? That's the real question - the MS textbook platform strategy days of bundle, co-opt and extract are, I suspect, long over now.