If we take it as true that one of the following scenarios is true: BIV (or similar) is true and I do not have knowledge of an external world. 2. BIV (or similar) is false and I do have knowledge of the external world. If we require an internal justification then it is impossible to know whether (1) or (2) is true (as both appear identical to us), so we cannot have knowledge at all. However if reliabilism is true and our knowledge has to be from a reliable process then it just has to be the case that (2) is true for us to have knowledge. So according to reliabilism we can have knowledge of the world, but we can't show that that the process is reliable. Both views have downsides, but if we need to internally justify then knowledge is impossible, if an external reliable process is needed then knowledge is at least possible.
 Criticise:
The reliabilist says we cannot know we are not BIVS, but if we are not BIVs then we do have knowledge. This could be compared to having ÂŁ1,000,000 in a bank account we know nothing about it may be the case we are millionaires but this is meaningless unless we are aware of the ÂŁ1,000,000. In the same way to have knowledge of the external world, but not be able to justify that knowledge might be considered a meaningless claim.