Social Security is not always "supposed to run out in 50 years," it's supposed to run out in 2033. Those predictions have been moved up in recent years, because the Great Recession depressed payrolls (and thus payroll taxes) so much. Does anyone who ever talks about Social Security bother to look up what Social Security's own actuaries are saying? The OASDI Trustee's report is not obscure or hard to find. Here's the report from 10 years ago, predicting trust fund exhaustion in 2042, before the Great Recession happened. Here's the most recent projection from a few months ago, predicting the date as 2033. Once the trust fund is exhausted, current payroll taxes will only cover approximately 74% of program costs; the difference will have to be made up by either benefit cuts or increases in the payroll tax.
This isn't some Republican conspiracy theory. You will note that the second letter is signed by a number of Democratic Party appointees, including three of Obama's cabinet members. These projections are not controversial among people who study Social Security for a living, in the same way that the atmospheric effects of CO2 are not controversial among climate scientists. The controversy only arises about how to fix the problem - nobody wants to be the first to propose 25% benefit cuts or 25% payroll tax increases.