A Key Power of the President is to Coordinate the Execution of Existing Concrete Plans

I listened to the 80,000 Hours pod­cast with Tom Kalil, who spent 16 years as Deputy Direc­tor of the Office of Science and Tech­nol­ogy Policy at the White House. Kalil seems skil­led at eval­u­at­ing con­crete sci­en­tific plans to offer the pres­i­dent and find­ing the path of least re­sis­tance through gov­ern­ment to effect those plans, though he is not him­self some­one with deep tech­ni­cal un­der­stand­ing of any sin­gle do­main.

One key idea I took from the pod­cast was that his main use of the ex­ec­u­tive branch of gov­ern­ment is as a co­or­di­na­tion mechanism. I moved away from think­ing of the Pres­i­dent as an ex­pert who makes de­ci­sions like a CEO, and much more as an in­di­vi­d­ual with im­mense co­or­di­na­tion power try­ing his best to take any con­crete plans given to him and co­or­di­nate the coun­try around ex­e­cut­ing on them. That is, not some­one who comes up with plans, not some­one who ex­e­cutes on the plans, but some­one who co­or­di­nates peo­ple to ex­e­cute the con­crete plans that are wait­ing to be picked up and run with.

Below are rele­vant and very in­ter­est­ing quotes, fol­lowed by a few more up­dates I made listen­ing to the pod­cast.

Key Quotes

Robert Wiblin: So, do you think peo­ple un­der-ap­pre­ci­ate how much the ex­ec­u­tive branch can just do au­tonomously?
Tom Kalil: Yes. Yeah. Not only what it can do, but the Pres­i­dent’s abil­ity to con­vene.
[...]
Tom Kalil: One thing that I used to ask peo­ple is to imag­ine that you have a 15 minute meet­ing with the Pres­i­dent in the Oval Office and he says, “Rob, if you give me a good idea for”, pick your cause, “re­duc­ing ex­is­ten­tial risk, then I will call any­one on the planet. It can be a con­fer­ence call so there can be more than one per­son on the line. If it’s some­one from in­side the gov­ern­ment, that I can di­rect them to some­thing be­cause I’m their boss, and if it’s some­one out­side the gov­ern­ment then I can challenge them to do some­thing. So, you not only have to tell me, what is your idea, but in or­der to make your idea hap­pen, who would I call and what would I ask them to do?”
Tom Kalil: There are sev­eral rea­sons for this thought ex­per­i­ment. One is that if you work for the Pres­i­dent you have the abil­ity to send the Pres­i­dent a de­ci­sion memo and have him check the box that says yes. Over time that give you a sense of what psy­chol­o­gists call agency, a sense that many things that you see in the world around you are the re­sult of hu­man ac­tion or in­ac­tion, as op­posed to the laws of physics. That’s one thing, a more ex­pan­sive view of what do you think is po­ten­tially change­able. The sec­ond is, it’s sort of a ver­sion of the Ham­ming ques­tion, pre­sum­ably if you re­ally did have a meet­ing with the Pres­i­dent you’d use it to de­scribe an is­sue that you thought was re­ally im­por­tant as op­posed to a sec­ondary or third tier is­sue. The third is that many com­plex prob­lems can­not be solved by a sin­gle in­di­vi­d­ual or­ga­ni­za­tion, they re­quire coal­i­tions.
Tom Kalil: You can’t build a coal­i­tion if you can’t ar­tic­u­late, num­ber one, who are the mem­bers of the coal­i­tion, and num­ber two, what are the mu­tu­ally re­in­forc­ing steps that you would want them to take. That’s one thing that I talk about in the Policy En­trepreneur­ship, then I also talk about some­thing that peo­ple don’t ever re­ally ap­pre­ci­ate, which is that policy mak­ers do things with words. What do I mean by that? Well, think about when the priest says, “I now pro­nounce you man and wife”, he has changed the state of af­fairs by virtue of, A, him be­ing a priest, and B, him say­ing, “I now pro­nounce you man and wife”. Similarly, the way that a policy maker both frames and makes a de­ci­sion and im­ple­ments that de­ci­sion is through doc­u­ments. When the Pres­i­dent does it we call it an ex­ec­u­tive or­der or pres­i­den­tial mem­o­ran­dum, when a reg­u­la­tory agency does it we call it a rule, when the Congress does it we call it leg­is­la­tion. But in all in­stances it is a doc­u­ment that you are cre­at­ing or edit­ing, so part of the policy pro­cess is that you are able to figure out what’s the doc­u­ment or doc­u­ments that you need to cre­ate or edit and who is al­lowed to take that some­thing from be­ing a Word doc­u­ment that is on your screen to some­thing that has some force in the world?
Tom Kalil: I would see this all the time, some­thing would go from be­ing a Word doc­u­ment on my com­puter to be­ing a pres­i­den­tial ex­ec­u­tive or­der, it always seemed like this slightly mag­i­cal trans­for­ma­tion from a Word Doc to some­thing that is in­struct­ing rele­vant mem­bers of the Cabi­net to take some ac­tion.
Robert Wiblin: I guess this makes you more am­bi­tious, then you’re like, “What is the best thing, what is the best memo that I could write.
Tom Kalil: Yeah, ex­actly, yeah. But also you have to be able to ar­tic­u­late some co­her­ent re­la­tion­ship be­tween ends and means. I would … a lot of times some­one would come visit me and they would say, “my is­sue is im­por­tant”.
Tom Kalil: I’d say, “great, let’s say that I’m pre­pared to stipu­late that, what is that you want me to do?”, then they would look at me and they would say, “Well, you should make this a pri­or­ity”.
Tom Kalil: I’d say, “What would that look like?” Peo­ple were not able … they were able to tell you that their is­sue was im­por­tant and that they thought the Pres­i­dent should de­vote more time and en­ergy to it, but then when you said “Alright, what is it, liter­ally … let’s say we got the Pres­i­dent su­per in­ter­ested in this is­sue, what would they do?” They weren’t able to ar­tic­u­late that part.
Robert Wiblin: Yeah, I’m sym­pa­thetic to that ’cause there’s a lot of things that I think are very im­por­tant, but I’m also not sure what should be done. I sup­pose maybe it makes sense for peo­ple to think more about that once they’ve got­ten peo­ple to care about it, but at the same time, maybe it’s hard to get peo­ple to care about some­thing if you have no ac­tual con­crete steps that they can take, they’re like, “Well, I don’t know what to do”.
Tom Kalil: Yes. Yeah, be­cause that’s as­sum­ing that … be­cause what you’re say­ing is, I’ve re­ally thought about this is­sue a lot and I think it’s re­ally im­por­tant, but I don’t know what to do.”
Robert Wiblin: That’s a bad sign. So you should think about it.

Kalil also talks about in his role as Deputy Direc­tor of the Office of Science and Tech­nol­ogy Policy, he helped raise the staff count from 40 to over 100 dur­ing the Obama ad­minis­tra­tion. He just gets to hire peo­ple who are ex­cited about an idea and want to make it hap­pen, and then they make it hap­pen us­ing the co­or­di­na­tion power of the ex­ec­u­tive office. Here’s a promi­nent ex­am­ple:

Tom Kalil: Let me give you one ex­am­ple. A young woman emailed me and the sub­ject line of your email was, “Cass Sun­stein says I should work for you.”
Robert Wiblin: That’s a strong sub­ject line.
Tom Kalil: Good sub­ject line. So I did a lit­tle re­search on her. It turned out that she had been a child vi­o­lin prodigy with Itzhak Per­l­man, had won the ma­jor Yale un­der­grad­u­ate awards, was a Rhodes scholar, and was wrap­ping up a post-doc at Stan­ford in De­ci­sion Neu­ro­science. I went out on a limb and I de­cided to take a chance on her. Her name was Maya Shankar. I asked Maya, “What do you want to do?”
Tom Kalil: She said, “The UK has cre­ated this or­ga­ni­za­tion called the Be­hav­ioral In­sights Team, which is tak­ing these in­sights from peo­ple like Kah­ne­man and Tver­sky and Sun­stein and Thaler and us­ing them to in­form poli­cies and pro­grams. Th­ese are all US re­searchers. Why don’t we have some­thing like this?” She said, “I would like to cre­ate that.”
Tom Kalil: Sure enough, in her late twen­ties, she ar­rived with no money, cre­ated this new or­ga­ni­za­tion called the So­cial and Be­hav­ioral Sciences Team, re­cruited 20 be­hav­ioral sci­en­tists to the fed­eral gov­ern­ment, got them to launch 60 col­lab­o­ra­tions with fed­eral de­part­ments and agen­cies and got Pres­i­dent Obama to sign an ex­ec­u­tive or­der in­sti­tu­tion­al­iz­ing this new en­tity.
Tom Kalil: I think that’s pretty con­se­quen­tial for some­one in their late twen­ties to be able to ac­com­plish. That’s one thing I did, was to re­cruit peo­ple of that cal­iber and teach them how to get things done in the fed­eral gov­ern­ment be­cause the gov­ern­ment doesn’t come with an op­er­at­ing man­ual.

The next quote is about how the core goal of the office of sci­ence and tech­nol­ogy policy is to take the nec­es­sary steps to get the pri­vate sec­tor to build new tech:

Tom Kalil: One of the things I learned is that if the United States is be­hind in a tech­nol­ogy, it’s very difficult to try to re-es­tab­lish a lead­er­ship po­si­tion… We tried to do that in the area of tech­nolo­gies like flat panel dis­plays and we in­vested some money, but I don’t think a whole lot came out of it.
Tom Kalil: Once Korea and Ja­pan dom­i­nated the mar­ket for things like ac­tive ma­trix, liquid crys­tal dis­plays, then try­ing to get the United States back into that mar­ket is re­ally, re­ally hard, and might re­quire more money than the US is will­ing to put into it. Be­cause ob­vi­ously we be­lieve that the pri­mary role of gov­ern­ment is to cre­ate the right en­vi­ron­ment for the pri­vate sec­tor. It’s not to en­gage in this sort of heavy handed-top down in­dus­trial policy that you see a China en­gag­ing in, for ex­am­ple… we in­vested in this idea of flex­ible elec­tron­ics where the idea is – maybe you have a dis­play that’s a piece of pa­per that you can roll up and put into your pocket. And, if that’s an area where no one has es­tab­lished a clear lead­er­ship po­si­tion, that’s more likely to be effec­tive than say­ing, okay, we’re go­ing to duke it out in some mar­ket that we’ve kind of already lost.

This fi­nal quote is an ex­am­ple of the co­or­di­na­tion power of the Pres­i­dent.

Robert Wiblin: Hav­ing worked in White House for 16 years, you must have some in­ter­est­ing or funny sto­ries from your ex­pe­riences there, differ­ent per­haps than what peo­ple ex­pect? Can you share one of them?
Tom Kalil: Sure. This is a story that hap­pened in 1995 and 1996, so as I men­tioned, Vice Pres­i­dent Gore was re­ally in­ter­ested in this idea of the in­for­ma­tion su­per­high­way and one of his goals was, what if we could con­nect ev­ery class­room to the in­ter­net? So I would tell peo­ple about the vice pres­i­dent’s in­ter­est in this is­sue and some­one who I’d got­ten to know, John Gage, who was at a com­pany by the name of Sun Microsys­tems said, “Oh, I’ve got this idea called Net Day. The idea is, what if on a sin­gle day, tens of thou­sands of en­g­ineers showed up in schools all across Cal­ifor­nia and started the pro­cess of wiring Cal­ifor­nia class­rooms to the in­ter­net?” I said, “Well great.” He said, “You know, I’ve got a web page of what this would look like if it ac­tu­ally hap­pened.” So he emailed it to me and I gave it to the vice pres­i­dent and the vice pres­i­dent thought it was a done deal.
Tom Kalil: So at this point, it was just in the fev­ered imag­i­na­tion of John Gage. So the vice pres­i­dent has a weekly lunch with the pres­i­dent and so he said, “Mr. Pres­i­dent, we have Sun, we have Ap­ple, we have HP, we have IBM, we have Pa­cific Gas and Elec­tric and they have all agreed that they’re go­ing to wire thou­sands of class­rooms and schools all across Cal­ifor­nia.” The pres­i­dent was like, “Great. Let’s an­nounce it.” So it turned out that they were go­ing to be in the Bay Area any­way so they de­cided, we’re go­ing to an­nounce Net Day. So I called up John Gage and I said, “They’re go­ing to come out and an­nounce this.” So he and I spent the next week call­ing in ev­ery fa­vor that we had to get these CEOs to show up and an­nounce that they were for us. They were a lit­tle sketchy on the de­tails of what it was.
Tom Kalil: What John did was he de­vel­oped a web­site, which was a click­able map of Cal­ifor­nia, that al­lowed you to zoom all the way down to the street level, all 12,000 pub­lic and pri­vate K through 12 schools had their own home­page. You could in­di­cate your level of ex­per­tise from, “I am an ex­pe­rienced net­work en­g­ineer.” To, “I will bring coffee and donuts.” All the schools were color-coded red, yel­low and green de­pend­ing on how many vol­un­teers had signed up. So we could look at the map and figure out which com­mu­ni­ties were get­ting on­board and which needed some pos­i­tive re­in­force­ment.
Tom Kalil: So they an­nounced that not only were they sup­port­ing it but they were go­ing to come back and per­son­ally par­ti­ci­pate in it. So by the time they did, we ac­tu­ally had tens of thou­sands of peo­ple who had vol­un­teered, so it was this pos­i­tive, self-fulfilling prophecy be­cause they said, “Oh, there’s go­ing to be a Net Day.” In fact, there was a Net Day and tens of thou­sands of en­g­ineers showed up to wire the schools and many par­ents showed up to wire the schools, but they dis­cov­ered the win­dows were bro­ken and the bath­rooms didn’t work, so a lot of them got more en­gaged in the schools as a re­sult.
Tom Kalil: Many states de­cided they were go­ing to do this and en­tire coun­tries de­cided that they were go­ing to have a Net Day as well. So it was this ex­pe­rience, a cou­ple of things that I took away from it, one is that you could cre­ate this pos­i­tive self-fulfilling prophecy, even though that was a very nerve-wrack­ing pe­riod of time for me per­son­ally be­cause I’d com­mit­ted to the pres­i­dent and vice pres­i­dent to do some­thing-
Robert Wiblin: To an­nounce the thing.
Tom Kalil: … and an­nounce some­thing that didn’t re­ally ex­ist yet. Right?
Robert Wiblin: Yeah, sounds a lit­tle bit like an epi­sode of Veep.
Tom Kalil: Then it was sort of ap­ply­ing mas­sive par­allels to this prob­lem. So as op­posed to say­ing, “How are we go­ing to wire 10,000 schools?” The ques­tion was, how could you get ev­ery com­mu­nity to take re­spon­si­bil­ity for one school? So it was just very in­ter­est­ing of the ex­pe­rience that I had of go­ing from some­thing be­ing a com­plete fan­tasy to ac­tu­ally see­ing it hap­pen.
Robert Wiblin: So gov­ern­ment can get things done.
Tom Kalil: Yes.
Robert Wiblin: Just in some­times a pe­cu­liar man­ner.
Tom Kalil: Yes, ex­actly.

There is a lot more fas­ci­nat­ing dis­cus­sion in the in­ter­view, es­pe­cially Kalil’s com­ments on us­ing fi­nan­cial prizes to in­cen­tivise sci­ence+tech in ar­eas like ed­u­ca­tion and poverty.

Updates

My new model is that the Pres­i­dent’s in­ter­ac­tion with sci­ence is largely to take con­crete ideas float­ing around in the en­vi­ron­ment that are ready for their time, and push them over the edge into ac­tu­ally be­ing built by the US pri­vate sec­tor, or into ac­tu­ally sub­stan­tially in­form­ing gov­ern­ment policy. This is similar to the no­tion that sci­en­tific ideas come about when the en­vi­ron­ment is ready for them (New­ton and Leib­niz both dis­cov­er­ing calcu­lus at the same time). There are ex­e­cutable plans float­ing around in the ether, and the Pres­i­dent keeps get­ting handed them and sets them off. His de­part­ment is not an origi­na­tor of new ideas, it co­or­di­nates the ex­e­cu­tion of ex­ist­ing ones. (And this does seem ob­vi­ously the cor­rect marginal use of at­ten­tion from the Pres­i­dent. Com­pare 15 min­utes per pro­ject ver­sus spend­ing a week be­com­ing an ex­pert in one and then ex­e­cut­ing it him­self.)

I’ve up­dated pos­i­tively on the tractabil­ity of gain­ing in­fluence within the gov­ern­ment and be­ing able to use it on timescales of 4-8 years. (I ex­pect I will likely make a fur­ther up­date when I read the blog­posts of Do­minic Cum­mings re­gard­ing UK poli­tics, though not sure how strongly.) Over­all I think in­fluence in gov­ern­ment, if you’re am­bi­tious and well-con­nected and have a very con­crete vi­sion, is likely quite a real ac­tion one can take. I ex­pect that from the per­spec­tive of gov­ern­ment there is a lot of low hang­ing fruit to be picked.

I up­dated nega­tively on the use­ful­ness of in­ter­act­ing with this part of gov­ern­ment in the short-to-medium term. My sense is that the state of un­der­stand­ing of how trans­for­ma­tive AI will be built and what im­pact it will have on the world is suffi­ciently low re­s­olu­tion and con­fused that we have no pro­ject or policy recom­men­da­tions for the gov­ern­ment, and will not be able to offer any­thing un­til we see fur­ther work that helps con­cep­tu­al­ise this space. Listen­ing to the pod­cast tells me that if you get 15 min­utes to talk to the Pres­i­dent about x-risk to­day, you are wast­ing his time, be­cause we have no con­crete plan that needs ex­e­cut­ing if only could co­or­di­nate ma­jor AI tech com­pa­nies. We have no R&D pro­jects that need fund­ing. We have no nu­anced AI-de­vel­op­ment poli­cies for global pow­ers to agree to. I’m pretty sure that there are peo­ple in this com­mu­nity who can co­or­di­nate Elon Musk and Demis Hass­abis or whomever else, should we have an ac­tion­able plan, but the cur­rent state is that we have no plan to offer.