Would it be feasible to have 10-watt tritium atomic batteries that with shielding and thermocouple loss included and weighed no more than your typical phone battery to power devices for 10 years or so assuming you can make tritium production cheaper(perhaps as a byproduct of fusion reactors)?