Elon Musk On The Future Of Warfare (YouTube Video Transcript)

(00:00:00) Your YouTube transcript will appear here (00:00:00) we're really excited for you to help us (00:00:01) kick off our our um intellectual theme (00:00:03) human in the machine which is leadership (00:00:05) on the emerging Battlefield how do you (00:00:07) see Warfare transforming in the future I (00:00:09) mean the biggest effect I think by far (00:00:12) is AI and drones uh so the the next well (00:00:16) in fact the current war in Ukraine uh is (00:00:19) very much a drone War already uh it's (00:00:23) it's sort of a contest between Russia (00:00:24) trying to see who can deploy the most (00:00:26) number of drones um now if there uh a (00:00:30) major power War it's very much going to (00:00:32) be a drone War it's it's going to be (00:00:34) drones and Ai and you know it's a sort (00:00:37) of I mean I do worry about the (00:00:39) existential risk of AI um which is that (00:00:42) if you if you employ Ai and drones do (00:00:45) you do you go down this path where (00:00:47) eventually you get to Terminator you (00:00:49) know try to avoid that um that would be (00:00:52) good that would be good yeah um minimize (00:00:55) the Terminator risk but but I mean (00:00:57) essentially if you are when when you're (00:00:59) making military Dres you are making (00:01:01) Terminators um (00:01:04) and you're I think you're you'll (00:01:07) somewhat be forced into the giving (00:01:10) localized the Drone localized AI um (00:01:12) because if the AI is far away it can't (00:01:14) control as well as localized AI so what (00:01:17) do you mean by localized AI meaning it's (00:01:19) autonomous killing machine completely (00:01:21) autonomous well you give it the okay (00:01:24) yeah in a particular Arena and it's and (00:01:26) it goes with certain parameters yeah (00:01:29) hopefully do you think our you think (00:01:30) think our adversaries will have that (00:01:32) those same type of concerns or (00:01:34) limitations um well yeah I mean depends (00:01:38) on how much existential risk there is in (00:01:40) these wars uh so if it's if it's a a (00:01:44) regional War I think it'll be more (00:01:45) tempered if it's goes beyond a regional (00:01:47) War then it's or betol and you know and (00:01:51) then you you start deploying things that (00:01:53) you really wouldn't want to deploy so (00:01:54) hopefully that doesn't happen but you (00:01:56) would but you said and I would agree (00:01:58) that if you just look at the (00:01:59) contemporary conflicts that are taking (00:02:01) place you would agree that machines (00:02:03) aren't just disrupting Warfare they're (00:02:04) now Common Place Dron drones are going (00:02:07) to be (00:02:08) overwhelmingly the uh what matters in in (00:02:11) any in any between you know any powers (00:02:15) that have uh significant technology it's (00:02:18) it's so my personal belief is like that (00:02:21) you really it'll actually be my I think (00:02:24) probably too dangerous to have humans at (00:02:27) the front it's it's drones it's drones (00:02:29) at the front it's too dous don't miss (00:02:32) because of The lethality though it's too (00:02:33) dangerous to have humans in the front (00:02:35) yes I mean I mean if you seeing some of (00:02:37) the computer controled sniper rifles I (00:02:39) mean they just don't miss so you're in (00:02:41) the you're you're finding a machine (00:02:44) that's going to you know aim with Micron (00:02:47) level accuracy and uh never gets tired (00:02:50) so how do you think the United States (00:02:51) should be Levering technology to further (00:02:53) our National Defense well I think we (00:02:55) probably need to invest in drones um the (00:03:01) um the United States is strong in in (00:03:03) terms of Technology of the items but the (00:03:06) production rate is low so it's it's a (00:03:09) it's a small number of units relatively (00:03:11) speaking but basically I think there's (00:03:15) there's a there's a there's a production (00:03:17) rate issue like theate say like how fast (00:03:19) can you make drones say you say there's (00:03:21) a drone conflict the outcome of that (00:03:23) drone conflict will be um how many (00:03:24) drones does each side have in that (00:03:27) particular Skirmish times the kill ratio (00:03:30) so if you've got you can have a let's (00:03:32) say we would United States would have a (00:03:34) drone a set of drones that have a have a (00:03:36) high kill ratio uh but then uh the other (00:03:39) side has far more drones if you got two (00:03:42) to one kill ratio the other side has (00:03:43) four times as many drones you're still (00:03:45) going to lose do you think our (00:03:46) industrial base can scale to make the (00:03:48) volume of drones that you're you're just (00:03:50) you're you're you know talking about I (00:03:52) think that's that's going to be the the (00:03:54) the biggest challenge that that is uh it (00:03:57) can scale but it is not currently (00:03:59) scaling why would that be I think the (00:04:02) procurement is so I mean there we not (00:04:05) read a lot of military history and and (00:04:07) actually like the thing that I go to (00:04:10) sleep with is using an audio book on on (00:04:13) U on military history on of one kind or (00:04:15) another uh so I find the subject very (00:04:18) interesting um and one of the things (00:04:20) that tends to happen is that uh (00:04:23) countries pretty much are geared enough (00:04:25) to fight the last war not the next war (00:04:28) and it's it's hard to change um I you (00:04:31) look at the the uh uniforms at the start (00:04:34) of World War I and the tactics and (00:04:36) strategies they use to start of World (00:04:37) War I there were not significantly (00:04:39) different from the Napoleonic era um you (00:04:42) know the the I mean the French were (00:04:44) marching into war with brightly colored (00:04:47) uniforms look great um that's that's (00:04:50) that's not what you want to be you know (00:04:52) when some trying to Point again at you (00:04:53) don't want a great looking uniform you (00:04:54) want uniform that blends in um so (00:04:58) there's a tendency to be geing after to (00:05:00) fight the last war the last war the US (00:05:02) for is kind of the Cold War I guess (00:05:05) so uh it usually takes like some kind of (00:05:08) shock factor to adjust uh I I would (00:05:11) recommend adjusting now um and and (00:05:14) you're are seeing some startups like and (00:05:16) and a few others that are have different (00:05:20) mindset um but it's really it's it's (00:05:23) going to be can you make a lot of drones (00:05:25) and what's kill ratio that's what it (00:05:27) comes down to so there was recently a (00:05:29) report that said that president zalinski (00:05:31) said by February 2025 there'll be a (00:05:33) million drones produced by the Ukrainian (00:05:35) so it seems like it's doable and this (00:05:37) might be a process question and we'll (00:05:39) talk about process in a second but as (00:05:42) you were as you were just talking I was (00:05:43) thinking about and you said that you you (00:05:44) can't have humans at the front and so (00:05:47) you haven't created a company that's (00:05:48) solved aging yet have you no okay so 100 (00:05:51) years I I wonder whether we should solve (00:05:53) aging that's a great Point yeah I'd like (00:05:56) to wrap it up sometimes right well it's (00:05:58) like yeah how long do you went to prud (00:06:00) and Kim J on to love that's a great (00:06:02) Point yeah but let's say don't get (00:06:04) pushed let's say you go for 50 100 years (00:06:06) how do you envision this Evolution and I (00:06:09) think this might get to neurolink how do (00:06:10) you see this Evolution between the human (00:06:12) who maybe can't be at the front any (00:06:13) longer the technolog is at the front yet (00:06:16) keeping them C integrated and (00:06:18) synchronized like how does that how is (00:06:20) that going to work in your mind I mean (00:06:22) so Communications is essential like it (00:06:25) is actually very important to have (00:06:27) space-based communications that are it (00:06:29) cannot be intercepted which is which is (00:06:32) starlink what starlink offers starink is (00:06:34) starlink is the backbone of the (00:06:37) Ukrainian military Communication System (00:06:40) um because it it can't be blocked by the (00:06:43) Russians essentially it's the only thing (00:06:45) they can't be so on the front lines all (00:06:47) the fiber connections are cut the cell (00:06:50) towers are blown up and the the (00:06:52) geostationary satellite links are jammed (00:06:54) the only thing that isn't jammed is slic (00:06:56) it's a only thing and then GPS GPS also (00:06:59) jammed (00:07:01) GPS signal is very faint but stalling (00:07:03) can offer location capability as well so (00:07:07) it is a a strategic advantage that's (00:07:09) very significant um and and when when (00:07:11) you're trying to communicate with the (00:07:12) drones the drones need to like basically (00:07:16) they need to know where they are and and (00:07:18) they need to receive instructions so if (00:07:20) you don't have Communications in (00:07:21) positioning then the drones don't work (00:07:23) that that's that's that's quite (00:07:25) important it's essential but you find it (00:07:27) important that there's still that that (00:07:28) communication between the human and the (00:07:30) the machine or the drum yeah yes there's (00:07:34) like it's a different question of like (00:07:35) where are things like right now versus (00:07:37) where will things be in 10 years um but (00:07:40) I have to say I I do look at the future (00:07:42) with sub trepidation I I have to have (00:07:44) some deliberate suspens of disbelief to (00:07:47) sleep sometimes um because I I think the (00:07:50) we're headed into a pretty wild future (00:07:53) and I'm I'm naturally an optimistic (00:07:54) person so you know but but AI is going (00:07:57) to be so good um including localized AI (00:08:01) I mean at at current rates you'll have (00:08:04) you know some of that sort of rock level (00:08:07) AI probably that can be run on a drone (00:08:10) and so you could literally say you know (00:08:13) um this is the equipment that the Drone (00:08:15) needs to destroy go into that thing (00:08:17) It'll recognize what equipment needs to (00:08:19) be destroyed and take it out it's a lot (00:08:21) of your work with neurolink though is (00:08:23) because what you're saying is that AI is (00:08:24) going to quickly surpass at least in (00:08:28) your estimation the the human's ability (00:08:29) to control it yeah okay I mean I'd like (00:08:32) to say no but the answer is yes so how (00:08:35) first off how long until you think that (00:08:37) happens before the the the the AI has (00:08:40) has evolved to the point where and I you (00:08:43) know they the AIS can start working (00:08:45) together even relying on computers like (00:08:46) in a de hobbled way and therefore (00:08:49) surpasses the ability for the human to (00:08:52) to be able to influence um how it's (00:08:54) working well I think humans will be able (00:08:56) to influence how it's working for a long (00:08:58) time um this is an esoteric subject that (00:09:01) you really goes into pretty wild (00:09:04) speculation um I I think it's to to some (00:09:07) degree that the AIS I think will want (00:09:10) humans as a source of will uh so if you (00:09:13) think of how the human mind works (00:09:15) there's lumic system and the cortex your (00:09:19) sort of your kind of um base instincts (00:09:21) and the sort of thinking and planning (00:09:23) part of your brain but you also have a (00:09:24) tertiary layer already which is that all (00:09:27) of the electronics that you use your (00:09:29) your phones computers applications um so (00:09:32) you already sort of have three layers of (00:09:35) intelligence um but all of those the (00:09:38) cortex and the Machine intelligence your (00:09:41) sort of cybernetic third layer is trying (00:09:43) to make the limic system happy because (00:09:45) limic system is a source a source of (00:09:46) will so there's some you know it might (00:09:49) be that the AIS just want to make the (00:09:52) humans (00:09:53) happy um and part of what neur Link's (00:09:56) trying to do is improve the (00:09:57) communication bandwidth between the (00:09:59) Vortex and the digital temp layer CU our (00:10:02) bandwidth band output bandwidth of a (00:10:05) human is less than one bit per second (00:10:06) per day there 86,400 seconds in a day (00:10:09) you don't outward 86,400 tokens so you (00:10:13) know it's but like the the number of (00:10:15) words that I can say at those Forum say (00:10:18) just like just look at it from an (00:10:19) information Theory standpoint how how (00:10:21) much information am I able to convey not (00:10:23) that much because I can only say a pure (00:10:24) number of words and in order in order to (00:10:27) convey an idea I have to take take a (00:10:30) concept in my head I have to compress it (00:10:31) down to a small number of woods try to (00:10:34) aspirationally model how you would (00:10:36) decompress th those woods into Concept (00:10:38) in your own mind that's communication so (00:10:41) your brain is doing a lot of compression (00:10:42) decompression and and and then has a (00:10:44) very small output bandwidth neuralink (00:10:47) can increase that bandwidth by several (00:10:50) orders of magnitude and uh and also you (00:10:53) don't have to spend as much time (00:10:54) compressing ports into a small number of (00:10:56) words you can do conceptual telepathy (00:11:00) that is the idea behind your link it's (00:11:02) so it is intended to be a mitigation (00:11:05) against AI existential risk you talk (00:11:07) about alignment can you explain what you (00:11:10) mean by alignment to to help everyone (00:11:13) understand yeah just is is AI going to (00:11:16) do things that make civilization better (00:11:19) make people happy or will it be contrary (00:11:22) to uh Humanity will it will it Foster (00:11:26) Humanity or or not will be against (00:11:28) humanity so obviously we want an that (00:11:30) will poster Humanity I think in (00:11:32) developing an AI to post Humanity (00:11:35) because I've thought about AI safety for (00:11:36) a long time um I think I've had probably (00:11:39) a thousand hours of discussion about (00:11:41) this um and my ultimate conclusion is (00:11:44) that the best course for AI safety is to (00:11:46) have an AI that is maximally truth (00:11:49) seeking uh and also curious and if you (00:11:52) have both of those things I think it (00:11:53) will Foster it will naturally Foster (00:11:56) Humanity because people want to see how (00:11:58) Humanity develops (00:12:00) humanity is more interesting than not (00:12:02) Humanity um you know I like Mars I'm a (00:12:05) big fan of Mars obviously um and and I (00:12:08) think we should become a multi-planet (00:12:10) civilization um I think that's that's (00:12:13) very important the purpose of SpaceX uh (00:12:16) is is to make life multiplanetary that's (00:12:19) the reason I created the company and and (00:12:23) that's the reason that we have the (00:12:24) Starship development in South Texas that (00:12:26) rocket is far far too big for just (00:12:28) satellites um it's it's a intended to (00:12:32) establish Life on Mars not just you know (00:12:35) send astronauts there briefly but to (00:12:36) build a city on Mars and that's (00:12:38) ultimately self- sustaining um so but (00:12:42) getting back to AI the if if you've got (00:12:45) a true seeking AI that is maximally (00:12:47) curious my neural net my biological (00:12:50) neural net says that that's going to be (00:12:51) the the the safest outcome um because (00:12:55) like see while like Mars you you could (00:12:58) Mars is not as interesting as us because (00:13:00) there's no human civilization there um (00:13:02) or thought of another way if you if (00:13:04) you're if you want to render Mars (00:13:06) rendering Mars is pretty easy it's (00:13:08) basically Red Red Rock on a look kind of (00:13:10) like some parts of Arizona you know as (00:13:13) not a lot of people um it's it's just (00:13:17) it's that's it's easy to render it's (00:13:18) render like to Mars but but rendering (00:13:20) human civilization much more much harder (00:13:22) much more complex much more interesting (00:13:25) and I so I think a curious truth seeking (00:13:28) AI would plus to humanity and want to (00:13:29) see where it goes that relies on (00:13:32) requires trust between the human and the (00:13:34) Machine and that's where I want to ask (00:13:36) you a question on this so the Army (00:13:39) leaders in the Army are um no strangers (00:13:41) to implementing new technologies um (00:13:43) think about how GPS for example (00:13:45) transform navigation it it'd be unheard (00:13:47) of not to use GPS uh today um but when I (00:13:51) was a lieutenant no one used GPS so (00:13:53) recently I was watching this incredibly (00:13:55) important and realistic documentary um (00:13:58) called Top Gun mavrick (00:14:00) yeah and in it I learned it's really (00:14:03) good it's really good it's a it's I mean (00:14:06) if you you don't want to think about the (00:14:07) plot too closely but it's a great movie (00:14:09) it's a fantastic movie I learned that (00:14:11) Tom Cruz is actually not an actor he's (00:14:13) like a pilot apparently but um he taught (00:14:16) me something really important in it he (00:14:18) says it's about the pilot not the plane (00:14:20) and right that's right before he defeats (00:14:22) a fifth generation fighter with a 19 14 (00:14:25) right yeah yeah so I just go with a B (00:14:28) plan yeah there you go Tom Cruz could do (00:14:30) it s with camel in it you know it's it's (00:14:34) a bit of a cynicism or a cynical view of (00:14:36) the need for technology it's like hey (00:14:39) technology is Superfluous humans can do (00:14:41) it but we know that's I don't question (00:14:43) Tom Cruz lot I don't ever question Tom (00:14:46) Cruz no I'm just kidding but I guess the (00:14:48) question is how do we get humans to be (00:14:49) able to trust the machines because there (00:14:51) is a lot of stories for example uh we (00:14:53) just recently had a conversation where (00:14:56) uh where a pilot aache Pilots were given (00:14:58) new technology and they were like we're (00:14:59) not going to use it because we don't (00:15:00) really trust it use it and so how do you (00:15:03) get the how do you when new technolog is (00:15:05) implemented we have to be able to trust (00:15:06) it especially if it's going to be the (00:15:08) Difference Maker to to win so how do we (00:15:09) do that how do we build the trust (00:15:11) between the human and the Machine well I (00:15:13) I don't I don't I think we we shouldn't (00:15:15) just automatically trust these things um (00:15:17) I mean I think you want to test it out (00:15:20) um you do a lot of testing and uh see (00:15:23) how it actually works in a conflict at (00:15:25) small scale and then scale it up if it's (00:15:28) if it's effective (00:15:30) um but uh yeah I mean I have to say like (00:15:33) I I'm not sure for example that there is (00:15:35) a um fortunately this is not an Air (00:15:38) Force Gathering but there's there's not (00:15:41) I'm not sure there's a lot of room (00:15:42) opportunity for fighter pilots um (00:15:45) because I think the if you've got a (00:15:47) drone swarm coming at you uh what what (00:15:50) the the pilots of liability in the fir (00:15:52) plane be honest uh so you know if if if (00:15:56) you say you compare u a drone for us as (00:16:00) a fighter plane how easy is it to make a (00:16:02) drone you you can it's an order of (00:16:04) magnitude maybe at least at least 10 (00:16:07) maybe 100 times easier to make the Drone (00:16:09) and you can afford to sacrifice the (00:16:10) drones whereas the pilots you don't want (00:16:12) to sacrifice the pilots so my guess is (00:16:15) that actually the the age of of human (00:16:18) palleted fighter aircraft is coming to (00:16:21) an end if that's the case uh then (00:16:23) there's a there's a a question that is (00:16:26) oftentimes debated and law and and (00:16:28) ethics debates about Killer Robots yes (00:16:32) and um really are these things that that (00:16:35) should we be willing to lean so forward (00:16:37) with the technology that we start to (00:16:39) supplant the the human pilot with the (00:16:41) technology and and where does that go um (00:16:44) and so what are your thoughts as we talk (00:16:46) about technology replacing humans on the (00:16:48) battlefield well I just what I'm saying (00:16:50) is that the the the at the front of the (00:16:53) battle lines it's going to be just (00:16:54) drones and any humans caught in the (00:16:57) crossfire are going to get it's killed (00:17:00) so it's it's it's a it's then then it's (00:17:02) irrelevant it's just going to be the (00:17:03) it's just going to be the way military (00:17:04) operations take place there isn't going (00:17:06) to be if you make the choice to be there (00:17:08) then you're you're at a significant (00:17:09) disadvantage yeah I mean I think it's (00:17:12) just think like you R drones that that (00:17:15) you know are constantly scanning they're (00:17:17) scanning in infrared scanning (00:17:20) invisible there's thous thousands of (00:17:22) them are t, you mention a million that (00:17:25) Ukraine's going to make got you got a (00:17:26) million drones coming at you yeah do you (00:17:28) do you want to be on (00:17:30) there with the try try to take out (00:17:32) drones of assault rifle it's like not (00:17:34) going to be a good situation um I mean I (00:17:37) think there there is there is something (00:17:39) where um if you go fully uh analog where (00:17:44) if you if you're if you can do um sort (00:17:47) of an Emi like electromagnetic explosion (00:17:51) of some kind that could trick it take (00:17:53) out Electronics but then you then your (00:17:54) electronics are going to go too so (00:17:56) you're going to go either fully analog (00:17:58) or fully digital um so I think that (00:18:00) there actually would be a role for a (00:18:02) fighter plane if it was fully analog and (00:18:04) had mechanical controls um because then (00:18:07) you could do an em sort of an EMF blast (00:18:10) take out the drones and the analog fight (00:18:13) I mean that could be anotherr Tom Cruz (00:18:15) mie maybe I don't know um you know um it (00:18:19) just goes a fully analog aircraft and (00:18:21) all the drones fall out of the sky (00:18:22) because of an EMF bomb how do you reply (00:18:24) to those in uh say industry that would (00:18:27) say I don't we don't want to contribute (00:18:29) to the the development of technology (00:18:30) that can be used by the Department of (00:18:32) Defense like basically we need to build (00:18:33) we need to build trust with with the (00:18:35) industrial base and with Society maybe (00:18:37) something we're doing what I mean how do (00:18:38) we how do we do that well I'm I'm very (00:18:41) prom military just so to be clear it's (00:18:43) good your audience will like that (00:18:47) yes (00:18:49) so so but I think what what what uh well (00:18:54) if if there's a significant conflict the (00:18:57) US industrial base will switch quickly (00:18:59) to military act military production just (00:19:01) did in World War II um is it you know is (00:19:04) it quick enough I don't know but that's (00:19:05) what probably happen um but yeah Ai and (00:19:09) Rance that's that's the future of (00:19:11) warfare and I mean tell me if I'm (00:19:14) missing something here where you where (00:19:16) do you see where do you see the domain (00:19:17) of space space uh yeah space is I mean (00:19:21) space is the ultimate High Ground so it (00:19:24) really goes Spacey big um real big so (00:19:28) well yeah you ever see like Earth to (00:19:30) scale with the sun and the you know it's (00:19:32) like wow we're just like a tiny little (00:19:34) dust mode you know floating around space (00:19:37) that's but space is becoming (00:19:39) increasingly militarized and so how did (00:19:41) you see that especially as it relates to (00:19:43) land Warfare like what's your thoughts (00:19:45) on the space domain as it relates to (00:19:47) land Warfare and and what are things (00:19:49) that we should be doing to to start to (00:19:52) gain those advantages that are necessary (00:19:53) well I mentioned I mentioned the (00:19:55) space-based communications is is (00:19:57) critical like if you can't communicate (00:19:59) you don't know what's going on can't (00:20:00) receive orders you can't report (00:20:03) information um and uh and whether it's a (00:20:06) human or a drone they need communication (00:20:07) so you got to have communications any (00:20:09) groundbased Communications like (00:20:11) fiberoptic fiber optic cables and uh (00:20:15) cell phone towers will be destroyed so (00:20:17) it's it's basically only all you've got (00:20:19) are basically analog radio radios and um (00:20:23) before any kind of data Communications (00:20:24) it's space based um and then while GPS (00:20:28) has been effective for a long time GPS (00:20:29) jamming at this point is pretty easy (00:20:31) because the GPS signal is is it's a weak (00:20:34) signal so it's easy to J do GPS jamming (00:20:38) um so having sort of a Next Generation A (00:20:40) system that can provide positioning is (00:20:43) is going to be very important um space (00:20:46) can also probably offer you know the the (00:20:48) ultimate weapons where you just have um (00:20:51) you know tungsten cannibals from allit (00:20:53) um how about offensive weapons in space (00:20:55) do you see those that's what I mean by (00:20:57) rods from God yeah so if you have like (00:21:00) you know CL they talked about this in in (00:21:02) the star war program in the80s but this (00:21:04) is certainly something we can't read (00:21:05) done which is you have just kinetic (00:21:07) weapons from space or space-based lasers (00:21:10) um soling system technically does have (00:21:13) lasers um but they're low power lasers (00:21:17) for now yeah so let me let me ask you (00:21:19) about back to this question about um (00:21:22) process process so so I like mil period (00:21:26) history also yeah uh so in um so in I (00:21:30) like all disciplines of the West Point (00:21:31) by the way I love all of so in um 149 BC (00:21:36) there was the third Punic Wars ongoing (00:21:38) and the Roman Legions are outside (00:21:39) Carthage and uh they lay Siege to (00:21:42) Carthage and it's not going very well (00:21:45) the pro councils that are in charge are (00:21:48) passive uh risk averse um and they're (00:21:52) losing and there's a young guy who's (00:21:54) from the famous skipio line of pro (00:21:56) counsels uh and it is it is skipio (00:22:00) amanas um who is the grandson the (00:22:03) adopted grandson skipio africanas and so (00:22:06) skipio is the only one who's doing (00:22:08) something and so KO the Elder is sitting (00:22:10) in the in the Senate and he says this he (00:22:13) says he alone still thinks the others (00:22:16) flit about in the shadows and his (00:22:18) basically argument was I want skipio in (00:22:21) charge and the problem was skipio was (00:22:23) too young you had to be 42 to be a pro (00:22:24) counsel and so K's like I don't care (00:22:27) he's the right guy and then was skipio (00:22:29) do he goes in he puts juice in's (00:22:31) Innovative and and they and obviously we (00:22:33) know have the third Punic Wars because (00:22:35) we know about Rome not Carthage right so (00:22:37) um what K was getting at is this need (00:22:39) for Innovative and creative and (00:22:41) entrepreneurial leaders right that's (00:22:43) what is necessary and so processes were (00:22:45) only as good as those who lead it and so (00:22:47) what are the traits you look for and (00:22:49) those who lead your various businesses (00:22:51) and Enterprises and I'm very much in (00:22:53) technology so for me uh if somebody is (00:22:56) going to lead something in technology (00:22:59) they must themselves be good at (00:23:00) technology meaning that if if they're (00:23:03) going to lead something that involves (00:23:06) complex engineering they must themselves (00:23:08) be good at at engineering they don't (00:23:10) necessarily need to be the best engineer (00:23:12) on the team but they need to be a they (00:23:14) need to be very competent in in their (00:23:16) field um so this is this is incredibly (00:23:20) important uh to to me if if if (00:23:23) somebody's leading a given engineering (00:23:25) field or engineering department and they (00:23:27) are not good at that that would be like (00:23:29) a Cavalry Captain who can't ride a horse (00:23:32) problem problem great leader in every (00:23:34) way except can't ride a horse um and (00:23:37) then you got to charge into battle and (00:23:39) Calvary Captain falls off the horse you (00:23:40) know it's not aspiring so uh Calvary (00:23:45) Captain must be able to ride a horse um (00:23:47) that's that's so that that that's (00:23:49) actually they they need to be the best (00:23:51) horse rider but they must be competent (00:23:53) in this regard otherwise they cannot (00:23:54) evaluate the talent of the team um and (00:23:57) they don't understand the technology (00:23:59) that's being developed this may seem (00:24:00) like a simple thing but it is often the (00:24:03) the case that this this is overlooked um (00:24:06) you know I don't want to pick on the CEO (00:24:08) boing but uh go to you know degree in (00:24:11) accounting or something um which I think (00:24:14) that's the you know you want to have (00:24:15) like a someone who knows how airplanes F (00:24:18) work running the airplane company I (00:24:21) guess I cross out my job that Boeing CEO (00:24:24) there can't do that I me you know it's (00:24:26) like it's like you want to you want to (00:24:28) not be the it's just if if you're (00:24:30) running an airplane company you should (00:24:31) know how airplane you should know how (00:24:33) airplanes work and how they fly and how (00:24:35) to design airplane um I think that's (00:24:38) pretty important um so I think it's (00:24:40) vital but how do you create Innovative (00:24:43) intuition in those that work for you I (00:24:45) mean you're you're you're famous (00:24:48) for uh trying to gain efficiencies (00:24:51) create create better processes pushing (00:24:54) to try to to try to um to gain those not (00:24:57) just efficiencies but Effectiveness so (00:24:58) how do you is it possible can you build (00:25:00) this Innovative intuition in a person (00:25:03) well I think it is possible to learn to (00:25:05) be Innovative you know a lot of times (00:25:07) for for any given thing uh you have to (00:25:10) say did you try this may sound sound (00:25:14) obvious uh but actually try like you (00:25:19) somebody might wonder well can I be (00:25:20) Innovative well have you tried try (00:25:23) thinking of interesting ideas I mean I I (00:25:25) do find a good source of of innovation (00:25:27) is if you read if you read about a whole (00:25:30) bunch of fields you can cross balize (00:25:32) ideas from one field into another and so (00:25:34) you can (00:25:35) synthesize t a SpaceX and Tesla the (00:25:39) automotive industry is very good at (00:25:41) manufacturing um it's a in terms of (00:25:44) manufacturing complex machines at volume (00:25:47) the automotive industry is the best the (00:25:50) now the rocket industry space industry (00:25:53) is very good at Advanced Materials and (00:25:55) making things very light um and uh and (00:25:58) so so taking Advanced Materials and mass (00:26:03) optimization uh Concepts from the space (00:26:05) industry applying it to automotive and (00:26:07) taking Automotive Mass manufacturing (00:26:10) techniques and applying it to space was (00:26:12) kind of like a superpower but when you (00:26:14) that's interesting is when you think (00:26:15) about it when you're talking about (00:26:16) innovating though and you said people (00:26:17) can try that means you have to be (00:26:19) willing to let them fail yes and so (00:26:21) where do you draw the line between (00:26:24) recklessness and being overly cautious (00:26:26) no if if you're not if you're not (00:26:28) failing at least some of the time you're (00:26:29) not trying hard enough uh you have to (00:26:31) fail some of the time um so you know (00:26:34) it's more like a batting average (00:26:35) somebody should have a good batting (00:26:37) average but nobody bats a thousand but (00:26:38) if somebody bats zero all the time I (00:26:40) mean okay you know you got to take them (00:26:41) off (00:26:43) um (00:26:45) so uh so I think I I do have this sort (00:26:49) of simple first principles algorithm (00:26:51) that I think could be quite helpful um (00:26:55) and I sort of say it as a monitor to (00:26:57) myself because I've made this mistake so (00:27:00) many times um so the first element is (00:27:04) for any given thing make make the make (00:27:07) the requirements less dumb um so so what (00:27:10) whatever problem you're solved make the (00:27:11) requirements less dumb and whoever gave (00:27:13) you those requirements even if they are (00:27:15) the smartest person in the world they're (00:27:16) still dumb so the so if if say like this (00:27:21) is where uh say military procurement it (00:27:24) goes wrong right at the outset with (00:27:26) excess requirements so you'll get sort (00:27:29) of this giant document of requirements (00:27:31) that actually should be like one page um (00:27:34) so step one make the requirements (00:27:36) simplifying just make requirements less (00:27:38) done um because if you don't make if you (00:27:40) don't do that as the first step then you (00:27:43) can get the right answer but to the (00:27:45) wrong question if the question's wrong (00:27:48) it's it doesn't matter so then then the (00:27:53) the step two is delete the part or (00:27:56) process step delete um and if you're not (00:27:59) putting in if if you're not adding back (00:28:01) 10% of what you deleted you not you (00:28:03) haven't deleted enough this again the (00:28:05) subance I think be very obvious but it's (00:28:08) very effective um but it goes the idea (00:28:11) is like if if you're not if you're not (00:28:14) if some of the ideas that you're doing (00:28:15) don't fail you're not trying hard enough (00:28:18) um and then only the third step is to (00:28:21) optimize the thing um and if I say like (00:28:25) what's one of the mistakes that I see (00:28:27) smart people making all the time (00:28:30) especially smart Engineers is optimizing (00:28:32) a thing that should not exist sounds (00:28:34) obvious you know um like you could try (00:28:37) to make let's have the world's best B (00:28:39) plane cloth B plane I'm like well (00:28:40) actually know we should have jet (00:28:42) airplanes instead you know um so we (00:28:45) should optimize the thing that should (00:28:46) not exist um and then step p is go (00:28:50) faster again it sounds really obvious (00:28:53) but people just don't try going faster (00:28:56) um and the the step would be to automate (00:29:00) something um but only automated once (00:29:03) you've done those those other four (00:29:05) things now the reason I I have this mtro (00:29:08) is because I've personally many times (00:29:11) automated something sped it up optimized (00:29:15) it and then deleted it and I'm like wait (00:29:17) I'm tired of going backwards here so um (00:29:21) if you run that that simple algorithm uh (00:29:23) in many Arenas of life you will be (00:29:26) shocked at how effective it is so shock (00:29:28) Le we are already running out of time um (00:29:32) so let me ask you this if you could (00:29:33) choose one attribute just one attribute (00:29:37) that' be critical for our future (00:29:38) officers uh to have to be successful (00:29:41) what would it be curiosity as long as (00:29:43) you're not not a cat but curiosity um (00:29:46) try to read as much as possible learn as (00:29:48) much as possible um and in in many (00:29:50) different fields um and apply critical (00:29:53) thinking to anything that you're told (00:29:57) thank you so I'd like to say on behalf (00:29:59) of Lieutenant General Gillan um and the (00:30:02) the the entire Academy um we're really (00:30:04) thankful that that you're here we're (00:30:06) really thankful you took the time to (00:30:08) help us uh celebrate the Excellence of (00:30:10) the faculty and the cadets um and really (00:30:14) sharing some wisdom with us because (00:30:15) we're really thinking about what do we (00:30:17) need to do to be successful uh because (00:30:20) we have a very important Mission which (00:30:21) is a no fail Mission which is we have to (00:30:24) fight and win and we're laser focused on (00:30:25) that yeah well I mean in in my view (00:30:29) or I think probably a lot of people's (00:30:30) views um you know uh America is like (00:30:36) like Atlas holding up the Free World and (00:30:38) you are the arms of Atlas thank you Mr M