A rubber wheel will produce significant resistance; you spend energy on deforming the rubber. Steel on steel (or aluminium on aluminium, whatever the flywheel is made of) is probably best to reduce losses. That is why trains have steel wheels and run on a steel track. As long as you have enough contact pressure, it won't slip.

With regard to the generator's RPM, the higher you can make it, the better it is in general; that produces the greatest change in flux density per unit of time (electromagnetic induction). So a relatively small wheel is probably best.

Also, using pedal power to generate electricity to run a motor to perform some task is only about 1/3rd as efficient as using the pedal power to directly perform the task.

Do you want something extremely simple and unbreakable? Get a hub dynamo. The downside is that it's more expensive and only a few times more efficient than what you were proposing.

If you want something really efficient, do read http://www.lowtechmagazine.com/2011/05/pedal-powered-farms-and-factories.html . (Low tech magazine in general is an extremely well researched and written publication.)

I would like to second mark's recommendation for low-tech magazine, and the related site no tech magazine.

Furthermore, doc has en excellent point about the inefficiency of using pedal power for electicity generation, as pointed out in detail in this lowtechmagazine article. Keep in mind that the human body is just not that powerful, so efficiency matters in this case. As you can see in the graph on the bottom of this page, a healthy human should be able to output 0.35 HP for about half an hour before being totally exhausted. That equals around 250W for half an hour, or 0.125 kWh. If you don't want to drop dead off the bike in half an hour, something like 100W would be more reasonable to expect. Currently, the most effiecient way to use pedal power is to use a bicycle for transport.

If you do choose to run a generator, the choice of a bicycle hub generator is not a very good one, though. Those are primarily designed for low drag when not in use, for the limited power needed for bicycle lights, and (in the case of Shimano and SRAM) for low cost in mass production. The best (and most expensive) bicycle hub generator are those from Schmidt Machinenbau in Germany. According to the documentation for their newest hub, it has an efficiency of 65% 'at 15 kph in a 700c wheel'. At this speed the output is only 3W. Its ouput plateaus at around 7W at 30 kph. Note that car alternators are not better rather worse. They are generally designed for higher output than a human can deliver (say 70A at 12V equalling 840W) You might expect to get 50-60% efficiency out of them at partial load, and much lower at the power levels a human could put in. To put that in perspective, a small hydropower generator can get up to 95% efficiency. Power station generators are around 99% efficient.

I've seen car alternators used, it's just a matter of getting the gearing right so it's not much drain on you, the flywheel should help you out too. I think it's a great idea to make energy (even if it's not a lot) while you ride since you would be doing it anyways. Pedal-A-Watt is a system that I've seen in use that may be worth looking at to get ideas on RPMs and DC motor sizes.

A disadvantage of car most car alternators is that they generally use a wound rotor (electromagnet) instead of permanent magnets. This construction is used because it enables the generator to put out a constant voltage at different speeds (by varying the strength of the magnetic field generated by the rotor), something that would not be possible with a permanent magnet rotor.

So you have to supply the alternator with current for the rotor to generate a magnetic field when starting it, otherwise it wouldn't do much. In a car, that current comes from the main battery. Once the motor is running, a fraction in the order of 10% of the generated current is needed to maintain that field. This is one of the reasons such an alternator will never be as efficient as one equipped with permanent magnets.

You could use this variability of the current through the rotor to control both power and voltage. But given the physics of inductance machines, the need to use part of the power to generate the magnetic field and the facts that the both the speed at which you can drive the alternator and a human's power output is a lot below the design point of a car alternator I suspect the efficiency to be very low.

A timely, interesting POV on pedal power elctricity: http://wattsupwiththat.com/2013/01/02/the-cost-in-human-energy/#more-76715
Summary: a human can steadily pedal out 100W/hr for, say, a ten hr day. He'd generate 1 kW-hr (average US consumption ~30kW/day), saving about 13 cents in energy costs.

What's the food bill to keep him going all day?

I greatly appreciate the information here. Needing to exercise more while having lots of online research to do, I would appreciate any suggestions you may have for building a device to allow me to sit in a chair and generate enough electricity to run my laptop. I don't know what is involved in making the connections, but am willing to read. I do not want to use an uncomfortable bicycle seat. If anyone has done something like this, let me know. Many thanks.

This site has great information.

Start with an old time manual treadle sewing machine. Hook a generator to its belt.