This essay is a continuation of “Peter Thiel, Palantir, and the Antichrist Problem.”
Nicole Flynn | September 2026
From evil to competence to Faust
In his September 2026 conversation with Mathias Döpfner, Peter Thiel moves from evil to competence, from competence to Faust, and then back to a qualification: it is better to be good than evil. A short excerpt circulating on X has brought the passage to a wider audience. Where the qualification is omitted, it belongs back in the account. What precedes it belongs there too.
A clarification can be sincere and useful to the speaker at the same time. It can correct a mistaken implication while preserving the proposition that prompted it. The passage therefore asks to be read twice: for the distinction Thiel makes, and for the question that distinction leaves open.
The first essay asked how infrastructure offered as protection could be distinguished from the enclosure it might enable. Here the question is closer to the sentence: what does a clarification resolve when the larger proposition remains in place?
Two meanings of good
A good person and a good instrument receive different kinds of praise. An instrument can perform beautifully in the service of something reprehensible; an admirable person can perform badly. Thiel brings the two uses into proximity. Evil receives the limited compliment of competence: at least it accomplishes something.
Faust supplies the comparison. In Thiel’s telling, Marlowe’s protagonist practices fraudulent or unsuccessful science and is damned; Goethe’s Faust acquires effective knowledge through morally dubious means and is redeemed. These are Thiel’s readings of the works. In the comparison he offers, accomplishment bears on the judgment of the actor.
Then comes the qualification: goodness is preferable to evil. The complaint about incompetence continues.
A listener can accept both propositions. Moral approval does not establish competence, and competence does not establish moral worth. The question arises at the point where accomplishment begins to count toward goodness. Something has worked. For whom, toward what, and at whose expense remain separate questions.
Nietzsche’s distinction between good/bad and good/evil offers a nearby philosophical reference, although his genealogy cannot be reduced to competence versus incompetence. The distinction needs no concealed source to matter here. Thiel makes it in public. The listener is left to consider the relation between the kinds of judgment it separates.
Eighteen days
On April 13, 2009, Thiel published “The Education of a Libertarian.” He declared that he no longer believed freedom and democracy were compatible, and identified women’s suffrage and expanding welfare constituencies as obstacles to capitalist democracy. The essay proposed technological frontiers beyond existing politics: cyberspace, seasteading, outer space.
On May 1, he published “Your Suffrage Isn’t in Danger. Your Other Rights Are.” It opens: “I had hoped my essay on the limits of politics would provoke reactions, and I was not disappointed.” He then rejected disenfranchisement, retained his doubts about voting, and again directed his efforts toward projects outside politics.
The provocation is acknowledged. What he had hoped it would accomplish is less fully specified. The clarification addresses an inference the original essay did not explicitly propose; it leaves the proposed technological exit in place. Deliberate provocation is on the record. A deliberate plan to mislead does not follow from it.
The second text clarifies his position. It also continues it. Someone reading only the first might infer a proposal to remove voting rights that the second expressly rejects. Someone reading only the second might miss the extent of the first essay’s departure from democratic politics. The right to vote remains. So does the proposed route around a political process he no longer expects to deliver the society he wants.
Seventeen years separate the essay and the interview. In one, the clarification arrives eighteen days later; in the other, it occurs within the passage. The issues differ, but the relation is comparable: a qualification addresses a narrower question than the preceding statement raises. Rejecting disenfranchisement does not resolve the proposed incompatibility of freedom and democracy. Preferring goodness to evil does not resolve what competence contributes to the judgment of either.
One reading is that the clarification makes a coherent position more exact. It rejects an implication Thiel does not hold and identifies how the retained project should be understood. On this account, preserving the project is what an accurate clarification should do.
The counter-account is that the qualification limits the charge that can be made against the speaker while leaving the consequential scope of the project unsettled. One can retain the vote yet lose influence over decisions made elsewhere. One can affirm goodness without explaining what would disqualify an effective undertaking from its service.
These are accounts of what the qualification does. Both can fit the words. A distinction can be sincere, accurately stated, and still insufficient for the decision to which it is applied. Their point of separation is whether the distinction supplies a limit that survives its passage from the explanation into the work.
A friend’s account
Eric Weinstein’s response offers an explanation from inside Thiel’s professional circle. Recalling a conversation from 2013, he publishes, as Thiel’s instruction to the foundation, this formulation:
“Good has two different antonyms. What is ‘bad work’? It could be seen as shoddy work. Or it could be seen as evil work. And we can’t do shoddy work. Why? Because evil that isn’t competent never gets to be evil. It’s just pathetic. So we have to be exceptionally competent at the good work which we do to make sure it is always confused for evil.”
Weinstein presents this as a recollection, not a transcript. The words reach us through him. Their significance here is the explanation they offer. Some useful work goes unfunded. Public approval can be mistaken. A foundation may choose work that markets will not fund and that contemporaries will call evil. None of those possibilities is difficult to recognize.
The last clause deserves attention: “to make sure it is always confused for evil.” As Weinstein recalls it, misunderstanding enters the stated aim rather than appearing only as a cost to be endured. The account leaves unclear whether this is provocative shorthand for pursuing unpopular work or an instruction to preserve its ambiguity. Either way, the explanation must leave room to distinguish an objection that misunderstands the work from one that identifies something wrong with it.
Now consider two undertakings, each competent, each unpopular, each supported by people convinced of its value. One deserves its eventual vindication. The other deserves the objections it attracts. Up to this point, the description fits both.
Their difference has to be found in the particulars: what is done, what follows, who benefits, who bears the cost, and what evidence could change the judgment. Disapproval cannot supply that difference by itself. Neither can the willingness to withstand it. Nor can a policy of maintaining the confusion, since that removes disapproval as evidence in either direction.
Weinstein’s recollection explains why a person might continue in the face of opposition. It does not tell an observer when that person should reconsider. An answer may exist within a particular undertaking. This account does not give it, and its absence here cannot establish its absence everywhere.
What would distinguish the readings
The two accounts make different demands on the record. The first asks whether the criticism depends on words or implications Thiel has expressly rejected. The second asks whether a question remains after those corrections have been fully accepted. Neither can be tested by shortening his answer.
For the 2009 pair, a charge that he explicitly proposed removing women’s votes fails against the texts. A question about how technological exit relates to democratic authority survives them. But an unanswered question is not yet evidence that the answer would be damaging. Exit might create a voluntary alternative; it might also allow decisions affecting others to move beyond their effective reach. A particular project would have to show which people can leave, which people remain subject to its decisions, and what recourse they retain.
The interview calls for the same discipline. Restoring his preference for goodness matters to any allegation that he simply endorses evil. It does not establish what criteria he uses to distinguish beneficial accomplishment from effective harm. Evidence that those criteria are stated and applied, especially when they constrain a favored undertaking, would weaken the concern that the distinction depends only on his assurance. Evidence that specific objections are repeatedly answered with declarations of good purpose, while the disputed effects go unaddressed, would strengthen it. The passages considered here do not supply that test.
An observer can therefore establish more about what the clarification retains than about why Thiel gives it. The unresolved part has a location. It lies between the account of the purpose and the evidence by which the undertaking can be judged.
Who bears the distinction
A clarification can change what may fairly be attributed to a speaker without changing the distribution of power around him. Part 1 identified a concrete instance of that distribution: Palantir’s founder voting structure. Its annual filing describes arrangements that preserve voting influence disproportionate to economic ownership and warns that the founders’ interests may diverge from those of other shareholders. Participation and control are separated in the company’s own account.
The structure can be defended as a means of protecting a long-term purpose against short-term pressure. The corresponding cost is reduced shareholder leverage over the people carrying that purpose forward. These are two consequences of the same arrangement. Whether the protection justifies the cost requires an assessment of the decisions it protects and the means of correction that remain.
This is a corporate governance example, not proof that Thiel’s political argument has been implemented through a share class. Shareholder rights are also different from the rights of people represented in a customer’s databases or affected by its decisions. The example establishes something narrower: the capacity to participate does not by itself specify the capacity to alter the course of an institution.
For a person affected by an institutional decision, that difference may be practical before it is philosophical. They may need to discover what information was used, correct a record, contest an inference, or obtain review from someone able to reverse the result. Whether those routes exist depends on the deployment, the institution, and the law. The interview supplies none of that information. A statement of good purpose cannot do its work for it.
The cost of a mistaken decision falls somewhere, whether or not its author understands the mistake. Establishing where it falls, and whether the person bearing it can obtain correction, requires a record of the undertaking. It cannot be settled by choosing the more persuasive description of the person behind it.
A philosophical qualification need not function as a statute. But when it is offered as reassurance about consequential power, the question changes: what does the distinction allow an affected person to require? The answer would show whether the stated limit operates beyond the speaker’s account of himself.
Thiel
Thiel repeatedly supplies a consequential proposition and the distinction through which he asks it to be understood. In 2009, the rejection of disenfranchisement accompanied a continuing argument for technological exit from politics. In the interview, the preference for goodness accompanies the distinction between moral judgment and accomplishment. Weinstein’s recollection supplies a reason to persist when public judgment is hostile. These are considered positions that can be examined together without deciding that their author has a concealed purpose.
The first essay asked how infrastructure offered as protection could be distinguished from the enclosure it might enable. Thiel’s acknowledgment that the restrainer can become entangled with what it restrains belongs to that inquiry. My further argument was that shared preventive capacities do not, by themselves, establish the moral difference. The new statements do not settle that institutional question. They return us to the relation between a purpose, the distinction offered in its defense, and the work carried out under it.
A shortened clip can misrepresent that distinction. The complete passage can leave it insufficiently specified. An admirer who restores the qualification has corrected the record; whether the larger concern is answered remains to be shown. A critic who regards every qualification as further proof of deception has made correction impossible. Neither response tells the person affected by the work what can be challenged or changed.
The first essay left Thiel’s theological identity unresolved. It did not leave his participation unresolved. He helped found Palantir and remains one of the founders whose voting influence the company describes. His public account of the dangers of concentrated power must be read beside that participation. The question is what his distinctions require of the institutions he helps sustain, including when honoring them would constrain a preferred course of action.
I cannot establish his ultimate purpose from these statements. That uncertainty does not suspend scrutiny of the decisions, the costs, or the authority to correct them. Evidence of a stated limit applied against a favored undertaking would matter. Evidence of specific objections repeatedly answered with declarations of purpose would matter too. Either must remain capable of changing the account presented here.
Thiel’s qualifications belong in the record because they change what can fairly be said of him. The people who bear the effects of an undertaking need something further: a means of testing its account of itself that does not depend on accepting it. His explanation is part of what they must be able to examine. It cannot be the condition of their right to do so.
Sources
[1] Peter Thiel, “The Education of a Libertarian,” Cato Unbound, April 13, 2009. Original essay.
[2] Peter Thiel, “Your Suffrage Isn’t in Danger. Your Other Rights Are,” Cato Unbound, May 1, 2009. Clarification.
[3] Peter Thiel with Mathias Döpfner, “Peter Thiel: The AI Crisis, Europe’s Decline & the Battle for America,” MD MEETS, September 2026. Full interview.
[4] Friedrich Nietzsche, On the Genealogy of Morals (1887), First Essay. The reference concerns different evaluative vocabularies; it does not identify competence with Nietzsche’s account of goodness or establish that Thiel was citing him.
[5] Eric Weinstein, X post dated September 26, 2026. Post. This is Weinstein’s retrospective account, not a contemporaneous transcript of Thiel’s remarks.
[6] Circulating excerpt and replies. The complete interview is linked in note 3.
[7] Nicole Flynn, “Peter Thiel, Palantir, and the Antichrist Problem,” Symfield. First essay.
[8] Palantir Technologies Inc., Form 10-K for the year ended December 31, 2025, filed February 17, 2026, discussion of founder voting arrangements and potential conflicts between founders and other stockholders. SEC filing.