Byrd station is not the GISP or Vostok CO2 data. The small figure below kinda shows how the Vostok record splices into the modern era. Essentially when CO2 rose at the end of the last glacial period, it went up to ~280 ppm, and that's where it stayed until the slight rise began in the mid-1700s/early 1800s.
"Splicing" (as with the temp hockey stick) is not valid science without showing natural variation in the older data (which you can't with ice cores since they average CO2 for a mininum of 30 years). The second problem is the Byrd station data is ludicrously cut off before it reached 285ppm +/- 10ppm about 10k years ago. Obviously a picture with an agenda.