Our type of intelligence doesn’t seem to be an end product of evolution, but rather an off-shoot. There’s no real reason to believe that it evolves on all planets where life has begun, after all it’s only happened here once in 3 billion years of evolution.

The other even more depressing thought, or maybe from another perspective amazing thought, is that life really is an improbable event and that it’s very very rare in the universe indeed. And of course if that’s true then this little blue world of ours is even more important and special.

They just don’t strike me as something you’d really want to build, for a whole variety of reasons.

So, to clarify – you’d have to wonder how intelligent ET is if you can’t decode the message. We’ve thought about how to encode a message in such a way that it’s reasonably easy to decode without any knowledge of who or what we are.

(OK, I’ve only read as far as the mid-twenties so far, so there might be something enlightening that I just haven’t reached yet.)

First, IIUC, nearly all natural sources of radio waves have broad bandwidths. Artificially-produced radio waves (except in perhaps the first 20 – 30 years of our use of radio) have narrow bandwidths. Certainly a suffciently narrow-band signal would be worth following up with additional observation.

Second, if a radio signal encodes information, there are only a very few ways of varying a radio signal. On/off keying, for instance (as in Morse code transmissions) would be pretty obvious, even if we couldn’t hope to decode the message’s content. Similarly, modulation of the frequency or amplitide of a signal would be obvious. A narrow-band signal with some kind of modulation on it would be a pretty good sign of ETI, but you’d have to question how intelligent ET is. After all, we’ve considered how we might encode a message to transmit to the stars, and we’ve concluded that the encoding would need to be pretty obvious (how about a repeating sequence of bits, where a positive frequency deviation represents one and a negative frequency deviation represents zero, in which the number of bits is the product of two primes – with this, we have encoded an image).

What are rocks made of? Metal and semi-metal oxides. Plus, only the crust of planets is rocky – the interior of the Earth is basically a big ball of Iron and Nickel. This is because early in Earth’s formation when everything was molten, the heavier elements sank down to the bottom and the lighter ones (oxygen, silicon, etc.) floated on top.

Plus, metal isn’t even the best construction material. Carbon fullerenes and graphene are superior building materials and carbon is relatively abundant in the Universe.

The main objection to a Dyson sphere is, as Chris said, the sheer *amount* of material you’d need. Considering the solar system, you’d need to break up every planet to get enough material for even a thin shell.

However, it’s quite possible that other star systems out there have a wealth of building materials. We’ve seen systems with a large number of huge planets. Plus, there’s always the possibility of towing in material from outside the star system.

Personally I think it’s likely that at least a few alien civilizations have built Dyson spheres, but probably not enough for us to have a high chance of detecting.

I think it is an acceptable margin of error because it’s within a single order of magnitude.

Having in mind that it was possible that the probability was around 0.001 or 0.0000001 (or even less… we had no idea after all), being sure that the probability is between 0.1 and 1 is pretty good indeed.

And yet SETI can’t find ET.

With all due respect to the efforts of the SETI people, right now SETI is akin to a colony of relatively intelligent ants trying to discover the existence of humans by trying to sniff the pheromones in the cubic metre of air around their nest.