Solar panels (such as my company makes) are an obvious solution, but they have a huge up-front capital cost. Diesel generators are less expensive per watt, but still require an up-front investment, and have ongoing maintenance and fuel costs. Biofuels (wood, vegetable oils, and so on) would work, but aren't an option when every bit of arable land is needed to grow food to keep from starving.
It would be nice if there was a dirt-cheap way for people to make electricity using materials readily available in their area. Aluminum beverage cans might be a solution.
Coca-Cola is sold in every country in the world, with penetration to nearly all remote communities. Packaging varies, but aluminum cans are used to some degree in nearly all markets. Worldwide, only a third of aluminum cans are recycled, with the rest being discarded. Therefore, even rural communities in developing countries are likely to have piles of old aluminum cans lying around.
It isn't hard to take an aluminum can, cut off the ends, and flatten it into a sheet.
Rub the sheet with sand to scratch through any varnish, paint, or epoxy coating. You can then snip the thin aluminum into conveniently-sized squares.
Wrap the squares in non-conductive, permeable material. A bit of cloth from an old polyester shirt would work.
Now, pour on a bit of salt water, or water dribbled through a bucket filled with ashes from a wood fire (sodium hydroxide and/or potassium hydroxide). Strip the insulation off some copper wire, and wrap that around the cloth.
The aluminum squares will start corroding to form a black gunk. You'll get a tiny amount of electricity flowing through the copper wire. The voltage will be tiny (0.7 V to 1.5 V per piece of aluminum, depending on whether you used salt or ashes in the water). You'll have to scrape the gunk off the squares frequently, and keep the copper wires clean. You'll have to wire the squares together in series for them to be of any use. Current flow will be limited by how fast oxygen from the air diffuses into the water.
Rough calculations show that the metal from one aluminum can could generate 1 amp for nearly a million seconds at 0.7 V. Theoretically. In real life, one aluminum can might run one LED lightbulb for a few days, or give a cellphone a partial charge. Even though your source materials are just trash, it might not be worth the investment in labour.
If I can find my multimeter and some LEDs, I'll see if this would actually work. Oxygen diffusion into the electrolyte is probably going to be the hard part.