Exploiting the Kepler transit data, we uncover a dramatic distinction in the prevalence of sub-Jovian companions, between systems that contain hot Jupiters (periods inward of 10 days) and those that host warm Jupiters (periods between 10 and 200 days). Hot Jupiters, with the singular exception of WASP-47b, do not have any detectable inner or outer planetary companions (with periods inward of 50 days and sizes down to 2R Earth ). Restricting ourselves to inner companions, our limits reach down to 1R Earth . In stark contrast, half of the warm Jupiters are closely flanked by small companions. Statistically, the companion fractions for hot and warm Jupiters are mutually exclusive, particularly in regard to inner companions.The high companion fraction of warm Jupiters also yields clues to their formation. The warm Jupiters that have close-by siblings should have low orbital eccentricities and low mutual inclinations. The orbital configurations of these systems are reminiscent of those of the low-mass, close-in planetary systems abundantly discovered by the Kepler mission. This, and other arguments, lead us to propose that these warm Jupiters are formed in-situ. There are indications that there may be a second population of warm Jupiters with different characteristics. In this picture, WASP-47b could be regarded as the extending tail of the in-situ warm Jupiters into the hot Jupiter region, and does not represent the generic formation route for hot Jupiters. Subject headings:1. FOREWORDS The origin of hot Jupiters (HJs, period inward of ∼ 10 days) has remained an unsolved issue. Although multiple scenarios have been proposed (disk migration, planet scattering, secular migration, etc.), none seem capable of satisfying all observational constraints. The recent discovery of two low-mass planetary companions (Becker et al. 2015) close to the hot Jupiter WASP-47b (Hellier et al. 2012) further obfuscates the picture. Motivated by the large population of low mass, closely-packed planets at small distances away from their host stars (Mayor et al. 2011;Howard et al. 2012;Borucki et al. 2011;Lissauer et al. 2011), and by the realization that some of them could have accumulated enough mass to undergo run-away gas accretion (Lee et al. 2014), Boley et al. (2016);Batygin et al. (2015) argue that WASP-47b, and possibly all hot Jupiters, were formed in-situ, instead of somehow transported inward. Only a tiny fraction of super-Earths need follow this path to be able to match the occurrence rate of hot Jupiters.While this seems a reasonable proposal for WASP-47b, could it explain the majority of hot Jupiters? To answer this, we focus on the following issue: is WASP-47b a generic hot Jupiter in terms of co-habiting with other planets? Currently, this question is best addressed by exploiting the Kepler data to look for small transiting bodies in systems hosting (either confirmed or candidate) There is a second goal to our paper: understanding the warm Jupiters (WJs). By this term we refer specifically to those giant planets orb...