Even when the law is perfectly clear based on stare decisis (precedent), the Court might choose a case because it is considering interpreting the Constitution differently than it had in the past (think Brown v. Board, Roe v. Wade, or more recently, Citizens United). As a result, a google search can only provide the legal arguments put forth by the appellant and appellee, as well as the state of the current law, but it cannot provide the “right” answer in a Supreme Court case. In fact, if such a google search existed, it would never allow Justices to find new rights in the Constitution, and if such a system had been in place in the last century, the Warren Court could not have found public school segregation to be unconstitutional, interracial marriage to be unconstitutional, etc. In that system, the judicial branch would lose the power of judicial review and would lose equal status with the other two branches of government.

As an aside, a google search absolutely can provide the right answer in a lower level court case, especially a trial court case, that must rely on existing precedent. In fact, most legal research today is done using databases that use google-like searches.

More generally, the role of a Supreme Court justice is to employ whatever methodology they feel is appropriate to interpret existing law and apply it to the facts of a real case. The methodology used (e.g. textualism, structure, original intent, original understanding, tradition, etc.) often results in a different outcome for a given set of facts. Depending on how skeptical you are of the Court, you might think that a justice chooses the methodology to employ for a given case to fit the outcome that they desire. Since there is no “correct” methodology, there is never a “correct” answer. And for this reason, machines would never be able to replace the Supreme Court.