If you read the note on Page 6 of the BLS (like it tells you to), you see that that 496,000 jobs that appeared out of nowhere in January is a statistical correction based on census estimates.
Didn't happen, they do it every year.

Adjustments to Population Estimates for the Household Survey

Effective with the data for January 2004, updated population controls have
been used in the household survey. Population controls for the household sur-
vey are developed by the U.S. Census Bureau. Each year, the Census Bureau
updates the controls to reflect new information and assumptions about the
growth of the population. The change in population reflected in the new con-
trols results primarily from adjustments to the estimates of net international
migration.

Official population and labor force estimates for December 2003 and earlier
months will not be revised. To assess the impact of the updated population
controls on trend growth, however, December 2003 estimates for selected data
series (not seasonally adjusted) were recalculated using the new controls, and
the differences from estimates based on the old controls are shown in table C.
The adjustments decreased the estimated size of the civilian noninstitutional
population by 560,000, of the civilian labor force by 437,000, and of employ-
ment by 409,000; the new population controls had a negligible impact on unem-
ployment rates and other percentage estimates. More detailed information on
the population adjustments and their effect on national labor force estimates
are available at http://www.bls.gov/cps/cps04adj.pdf on the Internet and also
will be published in the February 2004 issue of Employment and Earnings.