Reliabilism is a theory of knowledge that responds to the skeptic and to Gettier problems by asking if a knowledge claim was formed in a reliable way. Its focus is on method of belief formation. A belief such as "here is my hand" amounts to knowledge if it was formed in the proper lighting, without any hindrances to sight or other impediments. Here is a link to an article by Richard Feldman advocating such a view.
Like other contemporary theories of knowledge, reliabilism focuses on beliefs about the world based on sense data. I believe this is because reliabilists have already conceded skepticism about answers to basic questions, and they think that if humans have knowledge it must be about "ordinary things." However, the judgment about what is "ordinary" or "common" is itself reflective of a kind of epistemological value system, which as I said above reflects skepticism about basic things.
Problematically, reliabilism begs the question. To say that a method of belief formation is reliable is to say we know the mechanism that produces knowledge, and we can compare various instances to that mechanism. The skeptic will ask "how do you know this standard is accurate and produces knowledge?" If the reliabilist provides a further standard then this only steps the problem back and the skeptic will repeat his question. If the reliabilist appeals to common sense this again just pushes the problem back one step since common sense often errs. If the reliablist says it is probably accurate the skeptic will ask how this probability was determined, what is the denominator in the fraction of probability (3 out of 5 times this produces knowledge)? How was this determined?
The skeptic is asking if we are certain or if we are mistaken about our certainty. A pragmatic response by the reliabilist to the effect that no one doubts "here is my hand" misses the point. Epistemology is not first and primarily a practical problem. It is a problem of meaning. What do I mean when I say "here is my hand" and how do I know this?
What the skeptic is asking for is an authority. "By what authority do you say these things?" The contemporary move to common sense and ordinary language is not sufficient to avoid skepticism. Claims about my hand, or the table in front of me, are contextualized in a worldview (materialism, theism, dualism, idealism) and the skeptic will ask about that worldview framework.
If there is only an infinite regress of authorities to which we appeal then knowledge is not possible. If we end in our regress with an authority that could be mistaken then we end without certainty and in skepticism. What we are looking for is the highest authority which ends the regress and which cannot be doubted because it makes doubting possible. People rely on authorities such as tradition, testimony, scripture, common sense, intuition, science, and constructive reasoning. However, none of these is the highest authority; all of these can be (and are) questioned.
By way of contrast, I argue that the only authority that cannot be questioned is the laws of thought. The laws of thought, called reason, explain what thought is, therefore make thought possible, and hence cannot be questioned since any question assumes the laws of thought. The laws of thought are: identity (a is a), excluded middle (either 'a' or 'non-a') and non-contradiction (not both 'a' and 'non-a').
The reliabilist will ask how the laws of thought help us in knowing "here is my hand." I gave the response earlier. This belief is contextualized in a worldview, and so we must first examine the basic beliefs of the given worldview. If that worldview is materialism (as is often the case today among philosophers), then please see my post "something from nothing" for an example of how to examine the claims that either matter has always existed or that matter came into being by an uncaused event. If the materialist is wrong about this basic belief, then the truth value of "here is my hand" is the least of his problems.
Interestingly, we can conclude this post as reliabilists. Knowledge is the result of belief formation in a reliable way. Reliable belief formation is belief formation that begins with the most basic questions that can be asked (assumed by all other questions) and examines the various worldview answers for consistency. We are not seeking "belief forming mechanism aimed at truth," but rather "belief forming mechanism aimed at meaning" since when there is a contradiction meaning is lost. If a worldview assumes that matter is eternal, and upon examination we find this to be a contradiction, what we are finding is that the belief "matter is eternal" is without meaning.
And that is how I'd like to conclude. The skeptic should press us to ask not only how we know, but what do our beliefs mean. What is the worldview context of my supposedly "ordinary beliefs" and is this worldview meaningful?