World Government Summit 2017
NextY Combinator (2016)- Venue / interviewer: World Government Summit 2017, Dubai. An on-stage conversation between Mohammad Al Gergawi (UAE Minister of Cabinet Affairs) and Elon Musk. The YouTube upload, from the World Governments Summit channel, is titled “Mohammad Al Gergawi in a conversation with Elon Musk during WGS17”.
- Format: a ~35-minute on-stage Q&A. The topics run wide: the meaning of life and consciousness, multi-planetary civilization, the simulation argument, autonomous cars, “deep” AI and its regulation, tunnels under cities, automation and universal basic income, the human–machine merge, and a closing exchange on how he hires and what he’d tell the young.
- Date: February 13, 2017 (YouTube upload dated 2017-02-15).
- Trust tier: lower-trust-full-transcript (Tier 3). The text is YouTube’s machine caption track, not an official human transcript. Under the project’s Tier-3 rule, every quote has to be checked against the video before it can be cited. Where the caption wording is shaky, spreads across many cues, or the speaker is unclear, the line is paraphrased here instead of block-quoted.
- Quotes: from the official YouTube upload (there is no posted text transcript), each linked to its moment in the video.
- ⚠️ Attribution caveat (the biggest risk here): the captions carry no speaker labels, and two voices interleave throughout. The interviewer, Mohammad Al Gergawi, supplies the effusive intro (“Henry Ford, the Wright Brothers, Albert Einstein and Elon Musk,” “You are in rush, You want to go to places that nobody has been”) and all the questions (“What’s your life mission?”, “what is life for you?”, “Once you said you wanted to die on Mars. Why?”, “what three pieces of advice can you give them?”). Musk supplies the first-person answers (“Sure, first of all, thank you for having me…”). Only Musk’s answers are block-quoted; Al Gergawi’s intro, flattery, and questions are never put in Musk’s mouth. Speaker was decided by context (question versus answer) and checked against the video; anything even slightly uncertain is paraphrased.
⚠️ Tier-3 caption caveat. The caption track is fairly clean (punctuated, mixed case), but it’s still a machine transcript, not an official human one. Each block quote below is a short, distinctive Musk line whose caption rendering is internally clean and was checked against the video. Any caption slip is left exactly as the captions have it. Longer or broken passages, and any line whose wording drifts across cues, are paraphrased rather than dressed up as verbatim quotes.
Summary
For several of Musk’s recurring threads, this February-2017 conversation is one of the earliest clearly dated statements we have, and that’s the value of it. His core worldview is already fully formed in early 2017. On five separate subjects the talk fixes a date for where his thinking stood, somewhere between the June 2016 Code Conference and his later sittings.
The opening is the richest part. Asked for his “life mission,” Musk gives a tight version of his “understand the right questions to ask” and expand-the-scope-and-scale-of-consciousness creed. He’d say the same thing to Lex in 2021 and at DealBook in 2023; here it’s four years earlier, offered as the answer to a childhood “what’s the meaning of life?” question. From there he runs the rest of his stable arguments. Multi-planetary life as “life insurance for life collectively” and as a “sense of adventure”, a reason to get up in the morning (bright future). The Pong-to-photorealistic-games-to-are-we-in-one argument, almost word for word from the 2016 Code Conference. Full autonomy “in about 10 years”, with the elevator-operator image. The “deep”, general AI he calls “a dangerous situation”, with the advice that governments “keep a close eye” on it. The tunnels-under-cities congestion thesis, two months before the TED2017 reveal. The chain from automation to universal basic income to “how do people then have meaning?” The “we are already a cyborg” and bandwidth argument, with the “trillion bits per second” versus “10 bits per second” thumb arithmetic. And his physics-as-framework advice to the young, built on treating yourself as “always … to some degree wrong” and refusing wishful thinking.
The closing biographical exchanges are mostly event and character color: the China timing, the Washington-tunnel tease, his hiring question “tell me the story of your life,” the “I’m not sure I want to be me” aside. They’re summarized here rather than mined, since the focus is his mind.
Key quotes
The meaning of life — “the right questions to ask” and the scope of consciousness (Asking the right question, Curiosity and truth-seeking)
Asked his “life mission,” Musk traces it back to a childhood question about the meaning of life. His answer is the problem-finding creed the later sources keep restating, and this is the earliest statement of this exact frame that we have:
“And I came to the conclusion that what really matters is” ↗
“trying to understand the right questions to ask. And the more that we can increase the scope and scale of human consciousness, the better we’re able to ask these questions.” ↗
Years later this becomes the 2021 “foundation of my philosophy”, the 2023 DealBook “expand the scope and scale of consciousness … to figure out what questions to ask about the answer that is the universe”, and the 2023 #400 “philosophy of curiosity”. In February 2017 it’s already here almost word for word, and already pinned (as in those later versions) to a childhood “what’s the meaning of life?” crisis.
Multi-planetary life as “life insurance for life,” and a sense of adventure (Mars colonization, Humanity’s bright future)
He folds the meaning-of-life answer straight into the survival case, in its plainest insurance form:
“I think that being a multi-planetary species and being out there among the stars” ↗
“And, that’s one reason, kind of like life insurance for life collectively. Life as we know it.” ↗
But the part he says he finds “personally most motivating” isn’t the hedge. It’s the affirmative side, the bright future of adventure and inspiration:
“it creates a sense of adventure, and it makes people excited about the future.” ↗
“And there need to be reasons to get up in the morning.” ↗
“There’s got to be things that people find inspiring, and make life worth living.” ↗
This is the “reasons that you get up in the morning” idea, the same one the 2017 TED close uses two months later and the 2023 DealBook “what gives you hope?” turn returns to. It argues the multi-planetary case from inspiration, not only from asteroid-and-extinction insurance. He sets it against a two-futures contrast: one “forever confined to Earth until eventually something terrible happens,” the other “out there on many planets”. That’s the same binary the 2013 TED statement draws; the line runs across several cues, so it’s paraphrased here.
The simulation argument, restated (Simulation hypothesis)
Pressed on what life “is,” Musk runs the single extrapolation he’d given on stage at the 2016 Code Conference eight months earlier. Video games go from Pong to photorealism; project that forward at any rate of progress, and eventually they become indistinguishable from reality. Then comes the inference:
“then eventually those games will be indistinguishable from reality.” ↗
“how do we know that that didn’t happen in the past? And that we’re not in one of those games ourselves?” ↗
It’s the 2016 argument almost verbatim: the same Pong-to-photorealism setup, the same “any rate of improvement” assumption, the same conclusion that we’re probably inside one of the simulations. February 2017 sits between the June 2016 stage version and the 2018 Rogan “we are most likely in a simulation” compression.
Full autonomy “in about 10 years,” and the elevator analogy (Autonomous driving)
Here’s his timing call on self-driving, plus the everyday image he reaches for: the same elevator-operator analogy he uses again on Lex in 2019.
“it will be very unusual for cars to be built that are not fully autonomous.” ↗
“So, getting in a car will be like getting in an elevator.” ↗
He grounds it in the narrow-versus-general distinction the rest of his AI thinking depends on. Vehicle autonomy is “narrow AI … narrowly trying to achieve a certain function,” categorically unlike the “deep” AI he worries about (paraphrased; the distinction runs across several cues). This is his 2019 “a car that does not have autonomy would be about as useful as a horse” obsolescence thesis in its earliest dated form, with autonomy as the near-term default. He pitches it as a convenience here, not yet as a safety crusade.
“Deep” AI as “a dangerous situation,” and the regulation advice (AI existential risk)
Asked what technology will be most disruptive, he separates narrow AI from general AI and states the worry plainly:
“I think one of the most troubling questions is artificial intelligence.” ↗
“where you can have AI that is much smarter than the smartest human on Earth. This, I think, is a dangerous situation.” ↗
The framing he reaches for is the digital-superintelligence-as-alien image:
“Well, digital super intelligence will be like an alien.” ↗
Asked for advice to governments, his first item is the same call for outside oversight he makes from 2016 through 2019 (“a government agency that oversees anything related to AI”) and on into the 2023 sittings with Tucker Carlson and Bill Maher (“some regulatory body … so that they don’t cut corners”). Here it is in its February-2017 form:
“governments keep a close eye on artificial intelligence and make sure that it does not represent a danger to the public.” ↗
It’s the same warn-and-watch position throughout: the danger of “deep”, general AI, paired with a call for public-safety oversight. The decade behind his 2023 “Cassandra … for over a decade” line takes in exactly this period. His reasoning for the oversight, that researchers “can get so engrossed in their work” they miss the ramifications, is paraphrased here, since it runs across several cues.
Tunnels under cities — the congestion thesis, two months early (The Boring Company)
Teased about “building a tunnel under Washington D.C.,” he gives the 3D-tunnel argument he’d make publicly at TED two months later. It’s the earliest dated statement of it we have:
“the solution to urban congestion is a network of tunnels under cities.” ↗
“I mean tunnels that go many levels deep.” ↗
“Like, the deepest mines are deeper than the tallest buildings.” ↗
This is the TED2017 “deepest mines are much deeper than the tallest buildings are tall … alleviate any arbitrary level of urban congestion with a 3D tunnel network” thesis in seed form. Depth is effectively unbounded, so the solution to congestion has no ceiling either. He stresses going many levels deep, “20, 30, 40, 50”, and the challenge of building them fast, cheap, and safe (paraphrased across cues).
Automation, universal basic income, and “how do people then have meaning?” (Sustainable abundance)
This is the conversation’s most distinctive cluster, and the earliest statement of the whole chain among these sources: automation displaces labor, which forces universal basic income, which opens a crisis of meaning. He forecasts mass unemployment from autonomy and robotics, then names the policy conclusion:
“And I think ultimately will have to have some kind of universal basic income” ↗
“There will be fewer and fewer jobs that a robot cannot do better.” ↗
He is careful, as he stays throughout this thread in later years, to put it as a prediction rather than a preference:
“And I want to be clear, these are not things that I wish would happen.” ↗
“These are simply things that I think probably will happen.” ↗
The destination he reaches for is abundance, the 2025 master-plan word, already here in 2017:
“there will come abundance.” ↗
Then comes the move that matters most for his mind. He names the harder problem sitting under the economic one, the meaning question he’s still leaving open in 2024–2025:
“is how do people then have meaning?” ↗
The whole thread is the 2021 “work will become optional” + UBI and the 2025 “working will be optional … universal high income” argument, already fully formed in 2017. Automation is the cause, UBI the necessity, abundance the upside, and “do you feel useless?” the residual problem he can’t solve. He even puts a number on the job losses: “something like 12 to 15 per cent of the world force be unemployed” ↗.
“We are already a cyborg,” the digital ghost, and the bandwidth merge (Merging with AI, Human–AI symbiosis, Limbic–cortex model)
Asked where the next disruption leads, he restates the already-a-cyborg thesis. It’s his 2016 line in its February-2017 form:
“To some degree, we are already a cyborg.” ↗
He sketches the layered-mind picture: the limbic system, the cortex, and “your digital self as a third layer” (paraphrased; the layers run across several cues). Then he illustrates the digital layer’s permanence with an image the later sources drop:
“their digital ghost is still around.” ↗
The trajectory is a deeper fusion, bottlenecked by bandwidth. This is the bandwidth argument, carried by the arithmetic he’s best known for:
“a closer merger of biological intelligence and digital intelligence.” ↗
“A computer can communicate at a trillion bits per second, but your thumb can maybe do maybe 10 bits per second or a hundred” ↗
And his conclusion: a high-bandwidth brain interface as the thing that achieves symbiosis and, tellingly, addresses the AI control problem.
“So, some high bandwidth interface to the brain I think will be something that helps achieve symbiosis,” ↗
He closes the thought by naming what the symbiosis solves: “the control problem and the usefulness problem” (paraphrased; the phrase runs into the next cue). This is the 2016 “no longer a house cat” and “AI human symbiote … we are the AI, collectively” argument meeting the 2024 “less than one bit per second … like talking to a tree” arithmetic, bridged here in 2017. That puts it a year before the 2018 “tiny straw” version and seven years before the quantified #438 statement. The “trillion bits per second” against the thumb’s “10 bits per second” is the early version of the same throughput gap.
Physics as the framework, and “always … to some degree wrong” (First principles)
Asked for advice to the young, he names his thinking framework and the epistemic discipline under it:
“I’d recommend studying the thinking process around physics.” ↗
“the way of thinking in physics, it’s the best framework” ↗
The habit he says it teaches is to treat your own beliefs as provisional. Then he warns against the failure mode by name:
“always take the position that you are to some degree wrong, and your goal is to be less wrong over time.” ↗
“One of the biggest mistakes people generally make and I’m guilty of it too is wishful thinking.” ↗
“You ignore the real truth, because of what you want to be true.” ↗
And the corrective-feedback rule he comes back to again and again: ask friends for the bad news, because they won’t volunteer it.
“And solicit critical feedback, particularly from friends.” ↗
This is the physics-is-the-framework credo paired with the “less wrong over time” discipline and the refusal of wishful thinking. The same pairing shows up in the 2013 TED “solicit negative feedback, particularly from friends” line and the 2019 corrective-feedback-loop note. Here it’s his advice to the young, in February 2017. Earlier in the same interview he calls the corrective-feedback loop, and keeping it alive “even when people won’t … tell you exactly what you want to hear,” one of his biggest challenges (paraphrased, multi-cue).
The sustainable-energy mission, and the electricity-demand consequence (Sustainable-energy mission)
As the second of his three pieces of advice to governments, he restates the sustainable-transport-and-energy mission. From it he reasons to a concrete first-principles consequence: electrification triples electricity demand.
“which means that the demand for electricity will probably triple.” ↗
It’s the same mission as the 2013 TED “biggest problem we have to solve this century” line, surfacing in 2017 as the second of his three pieces of advice to officials. The electricity-demand arithmetic is the first-principles consequence he draws from it.
Connections (pages touched)
- Asking the right question — extended with the earliest dated statement of the “understand the right questions to ask” frame (February 2017), tied to a childhood meaning-of-life question, four years before the 2021 spoken version.
- Curiosity and truth-seeking — extended with the 2017 “increase the scope and scale of human consciousness … better able to ask these questions” creed, the earliest instance of the expand-consciousness philosophy among these sources, predating the 2021/2023 versions.
- Mars colonization — extended with the 2017 “life insurance for life collectively” insurance framing and the two-futures contrast.
- Humanity’s bright future — extended with the 2017 “sense of adventure” / “reasons to get up in the morning” / “make life worth living” inspiration case, the affirmative pole stated two months before TED2017.
- Simulation hypothesis — extended with the February-2017 restatement of the Pong → photorealistic-games → “are we in one of those games” argument, between the 2016 Code Conference version and the 2018 Rogan compression.
- Autonomous driving — extended with the 2017 “fully autonomous in about 10 years” timing call and the getting-in-a-car-like-an-elevator analogy.
- AI existential risk — extended with the 2017 “deep”/general AI “dangerous situation,” the digital-superintelligence-as-alien image, and the governments-keep-a-close-eye oversight advice.
- The Boring Company — extended with the February-2017 tunnels-under-cities / “deepest mines are deeper than the tallest buildings” congestion thesis, two months before the TED2017 public reveal.
- Sustainable abundance — extended with the earliest dated automation → universal-basic-income → abundance → “how do people then have meaning?” chain we have (2017), an early instance of the post-work thread the 2021/2025 sources develop.
- Merging with AI — extended with the 2017 “to some degree, we are already a cyborg,” the “digital ghost,” and the closer biological/digital merger framing.
- Human–AI symbiosis — extended with the 2017 bandwidth arithmetic (“trillion bits per second” vs the thumb’s “10 bits per second”) and the high-bandwidth-interface-as-symbiosis argument that names the control and usefulness problems.
- Limbic–cortex model — extended with the 2017 limbic / cortex / “digital self as a third layer” statement, an early form of the three-layer model.
- First principles — extended with the 2017 “thinking process around physics” framework, the “always … to some degree wrong … less wrong over time” discipline, the wishful-thinking trap, and the “solicit critical feedback, particularly from friends” rule.
- Sustainable-energy mission — extended with the 2017 sustainable-transport/energy advice and the electricity-demand-will-triple consequence.
- Elon Musk — extended with a “World Government Summit 2017: the early-2017 baseline” section threading the meaning/consciousness creed, the multi-planetary inspiration case, the simulation argument, the AI-risk + regulation advice, the automation/UBI/meaning chain, the cyborg/bandwidth merge, the physics-framework / anti-wishful-thinking advice, and his hiring / problem-solving heuristic (“tell me the story of your life” + verify real problem-solving by depth of detail) — his early-2017 baseline for these threads.