"Don't Teach Kids To Code!" -NVIDIA CEO

The Vaush Pit
28 Feb 202425:19

Summary

TLDREn este intenso monólogo, el locutor critica fervientemente las declaraciones del CEO de Nvidia, Jensen Huang, sobre la programación y la inteligencia artificial. Argumenta contra la idea de que la tecnología de IA debería reemplazar la necesidad de aprender a programar, destacando la complejidad y el arte detrás del código. El video aborda temas como la importancia del conocimiento técnico, la función y limitaciones de la IA, y el impacto de la tecnología en diversos campos. El locutor expresa su preocupación por el enfoque excesivo en la IA y los posibles efectos negativos en la innovación y el desarrollo de habilidades técnicas.

Takeaways

  • 😡 El CEO de Nvidia, Jensen Huang, sugiere que deberíamos dejar de enfocarnos en enseñar a los niños a programar, argumentando que la tecnología debe avanzar hasta el punto en que la programación sea innecesaria y accesible para todos.
  • 🤔 Esta postura es contraria a la opinión común de que aprender ciencias de la computación es fundamental, especialmente en los últimos 10 a 15 años.
  • 🚀 Huang promueve la idea de que la inteligencia artificial (IA) puede transformar a todas las personas en programadores al hacer que la interfaz de programación sea más humana y accesible.
  • 😠 El narrador expresa una fuerte desaprobación hacia la idea de Huang, argumentando que la programación es una habilidad crucial y comparándola con un arte que no puede ser completamente reemplazado por la IA.
  • 👨‍💻 Hay una crítica a la simplificación excesiva en la enseñanza de programación y tecnología, resaltando que las herramientas simplificadas no sustituyen la comprensión profunda y las habilidades técnicas.
  • 🖥️ Huang menciona que las herramientas de IA y sitios web como Squarespace han hecho que ciertos aspectos de la programación, como el desarrollo web, sean más accesibles para los no programadores.
  • 📈 El narrador cuestiona la sostenibilidad y la ética de reemplazar habilidades y trabajos humanos con soluciones de IA, especialmente en campos creativos y técnicos.
  • 🧠 Se discute la idea de que, aunque la IA puede ofrecer soluciones rápidas y accesibles, carece de la profundidad, adaptabilidad y creatividad de la inteligencia humana.
  • 🏫 Hay una defensa apasionada de la importancia de aprender y comprender la programación, no solo para el desarrollo profesional sino también para el crecimiento personal y la comprensión del mundo tecnológico.
  • 🌐 El debate subyacente trata sobre el equilibrio entre hacer que la tecnología sea accesible y mantener la necesidad de una educación profunda y habilidades en áreas como la programación y el diseño de sistemas.

Q & A

  • ¿Cuál es la opinión principal del orador sobre el CEO de Nvidia?

    -El orador está extremadamente enojado y en desacuerdo con las declaraciones del CEO de Nvidia, Jensen Huang, quien argumenta que nadie debería tener que aprender a programar debido a la inteligencia artificial.

  • ¿Por qué el orador se opone tanto a la idea de no tener que aprender a programar?

    -El orador cree que la codificación es un arte y una habilidad valiosa que requiere un profundo conocimiento y comprensión. Argumenta que la IA no puede reemplazar completamente el proceso creativo y la experiencia humana en la programación.

  • ¿Qué piensa el orador sobre el futuro si nadie aprende a programar?

    -El orador predice un futuro sombrío donde todos los juegos y software tendrán una calidad deficiente porque nadie sabrá programar correctamente. Cree que la IA no puede reemplazar completamente el conocimiento y las habilidades de los programadores humanos.

  • ¿Cuál es la opinión del orador sobre la inteligencia artificial generativa?

    -El orador odia profundamente la inteligencia artificial generativa y la considera antihumana. Cree que amenaza la creatividad y el arte, y está dispuesto a hacer cualquier sacrificio para detenerla.

  • ¿Qué piensa el orador sobre el enfoque de Nvidia en la IA?

    -El orador cree que Nvidia está promoviendo agresivamente la IA simplemente para impulsar su valoración en el mercado y atraer inversiones, en lugar de enfocarse en su negocio principal de GPUs y tecnología gráfica.

  • ¿Cuál es la opinión del orador sobre la precisión y confiabilidad de la IA en la codificación?

    -El orador duda mucho de la capacidad de la IA para codificar de manera precisa y confiable, especialmente en proyectos complejos. Cree que la IA no puede manejar adecuadamente el nivel de complejidad y los requisitos técnicos de la programación moderna.

  • ¿Qué piensa el orador sobre el uso de la IA en el descubrimiento de vacunas y medicamentos?

    -El orador parece estar más abierto al uso de la IA para procesar y analizar datos en áreas como el descubrimiento de vacunas y medicamentos, siempre y cuando no sea IA generativa.

  • ¿Cuál es la opinión del orador sobre los artículos y las valoraciones de mercado de Nvidia?

    -El orador cree que los artículos y las valoraciones de mercado exorbitantes de Nvidia son impulsados por la especulación y no reflejan el valor real de producción de la empresa.

  • ¿Qué piensa el orador sobre la accesibilidad y la simplificación de la programación?

    -El orador no se opone a hacer que la programación sea más accesible y simple, pero cree que siempre habrá un límite y la complejidad adicional seguirá siendo valiosa. No cree que la IA pueda reemplazar completamente la necesidad de aprender a programar.

  • ¿Cuál es la actitud general del orador hacia la modernidad y la tecnología?

    -El orador parece tener una actitud muy crítica y escéptica hacia ciertos aspectos de la modernidad y la tecnología, especialmente aquellos relacionados con la inteligencia artificial generativa y su impacto en la creatividad humana.

Outlines

00:00

😡 El CEO de Nvidia pide no enseñar a codificar y recibe críticas

El video muestra la furiosa reacción ante los comentarios de Jensen Huang, CEO de Nvidia, quien sugirió que ya no es vital que las personas aprendan a programar debido a los avances en la Inteligencia Artificial (IA). El narrador se opone rotundamente, argumentando que la programación es un arte complejo que la IA no puede reemplazar. Afirma que se necesitan habilidades de codificación para optimizar y personalizar tecnologías, y que la IA generativa actual produce resultados deficientes. Critica a Nvidia por centrarse en la IA para impulsar sus acciones.

05:00

🤖 Nvidia promueve la IA para atraer inversión

El video analiza cómo Nvidia está aprovechando el auge de la Inteligencia Artificial (IA) para impulsar su valor bursátil. Destaca que Nvidia se ha posicionado como líder mundial en computación de IA en su sitio web, a pesar de ser conocida principalmente por sus GPU para gaming. Sugiere que esto es una estrategia para atraer inversión especulativa, similar a burbujas tecnológicas anteriores como las criptomonedas y NFTs. Cuestiona las afirmaciones de Nvidia sobre las capacidades de la IA y advierte sobre los riesgos de confiar ciegamente en esta tecnología.

10:00

💻 La programación requiere habilidades humanas

El video argumenta que, si bien existen herramientas para facilitar el desarrollo de software, la programación requiere habilidades humanas esenciales que la Inteligencia Artificial (IA) aún no puede reemplazar. Destaca la importancia de comprender conceptos técnicos complejos, optimizar el código y permitir la personalización y modularidad. Afirma que los tutoriales en línea suelen enseñar métodos básicos e ineficientes, y que la IA generativa actual no puede producir código de alta calidad. Defiende la necesidad de aprender a programar en profundidad para desarrollar verdaderas habilidades.

15:02

🤑 Nvidia se enriquece con la burbuja especulativa de la IA

El video critica a Nvidia por capitalizar la burbuja especulativa en torno a la Inteligencia Artificial (IA). Menciona que Nvidia ha alcanzado una capitalización bursátil sin precedentes, impulsada por las expectativas de inversores sobre la IA, a pesar de no haber lanzado aún productos revolucionarios en este campo. Cuestiona las afirmaciones de Nvidia sobre las capacidades de la IA y sugiere que son principalmente estrategias de mercadotecnia. Expresa preocupación por el excesivo entusiasmo y la especulación en torno a la IA, que podría conducir a resultados decepcionantes.

20:03

⚗️ Usos legítimos de la IA en investigación científica

El video reconoce que existen aplicaciones legítimas de la Inteligencia Artificial (IA) en campos como la investigación de vacunas y el descubrimiento de fármacos. Estas aplicaciones se enfocan en el procesamiento de grandes cantidades de datos y la identificación de patrones y tendencias, en lugar de la generación de contenido. Se explica cómo la IA puede ayudar a analizar millones de datos y sugerir posibles direcciones de investigación, que luego son evaluadas y probadas por científicos humanos. Aunque se mantienen reservas sobre la IA generativa, se reconoce el potencial de la IA como herramienta de apoyo en la ciencia.

25:04

☪️ Odio visceral hacia la IA generativa

El video culmina con una expresión de odio visceral hacia la Inteligencia Artificial (IA) generativa, comparándola con el Anticristo. El narrador afirma que su aversión a esta tecnología supera incluso su rechazo a ideologías políticas extremas. Sugiere que estaría dispuesto a sacrificar instituciones científicas si eso significara detener el avance de la IA generativa. Esta postura radical refleja una profunda desconfianza y temor hacia las implicaciones de esta tecnología, que se percibe como una amenaza para la humanidad.

Mindmap

Keywords

💡Programación

La programación se refiere al proceso de escribir código de computadora para crear aplicaciones, software y sistemas. En el video, el orador expresa una opinión controvertida del CEO de Nvidia, quien argumenta que la inteligencia artificial eventualmente eliminará la necesidad de que las personas aprendan a programar. Sin embargo, el orador se opone enérgicamente a esta idea, argumentando que la programación es un arte complejo y que siempre será necesario el conocimiento humano para codificar sistemas avanzados que la IA no puede reemplazar.

💡Inteligencia Artificial (IA)

La inteligencia artificial se refiere a la capacidad de las máquinas para realizar tareas que normalmente requieren inteligencia humana, como el aprendizaje, la resolución de problemas y la toma de decisiones. En el video, la IA es el tema central de discusión, y el orador critica las afirmaciones del CEO de Nvidia sobre cómo la IA supuestamente reemplazará la necesidad de que las personas aprendan a programar. El orador argumenta que la IA tiene limitaciones significativas y que ciertas tareas complejas, como la codificación avanzada, todavía requieren habilidades y comprensión humanas.

💡Desinformación

La desinformación se refiere a la difusión de información falsa o engañosa, a menudo con la intención de engañar o confundir. En el video, el orador acusa al CEO de Nvidia de promover desinformación al sugerir que la IA puede reemplazar la necesidad de aprender a programar. El orador argumenta que estas afirmaciones son falsas y que Nvidia las está utilizando como una estrategia de marketing para impulsar su negocio de IA, a pesar de las limitaciones reales de la tecnología.

💡Capitalismo

El capitalismo es un sistema económico basado en la propiedad privada y la operación de empresas con fines de lucro. En el video, el orador critica el papel del capitalismo en la promoción de la IA por parte de Nvidia. Sugiere que la motivación principal de Nvidia es aumentar su valor de mercado y atraer inversiones, en lugar de promover una tecnología verdaderamente transformadora. El orador insinúa que el capitalismo impulsa a las empresas a exagerar las capacidades de la IA para atraer inversores, incluso si eso significa engañar al público.

💡Burbujas tecnológicas

Una burbuja tecnológica se refiere a un período de inversión excesiva y especulación en torno a una nueva tecnología, impulsada por el entusiasmo público y las expectativas infladas. En el video, el orador compara la ola actual de IA con burbujas tecnológicas anteriores, como las de las criptomonedas y los NFT. Sugiere que la IA podría ser la próxima burbuja tecnológica, donde las empresas como Nvidia inflan artificialmente las expectativas de la IA para atraer inversiones, a pesar de las limitaciones reales de la tecnología.

💡Creatividad

La creatividad se refiere a la capacidad humana para generar ideas y conceptos nuevos y originales. En el video, el orador argumenta que la IA no puede reemplazar la creatividad humana involucrada en la programación y el diseño de sistemas complejos. Sugiere que la programación es un arte que requiere habilidades creativas, como la iteración, el desarrollo y la comprensión de procesos complejos, habilidades que la IA actual no posee.

💡Optimización

La optimización se refiere al proceso de mejorar la eficiencia y el rendimiento de un sistema o proceso. En el video, el orador critica la sugerencia de que la IA puede optimizar automáticamente el código y los sistemas informáticos. Argumenta que la optimización real requiere un profundo conocimiento y habilidades humanas, ya que implica comprender y ajustar complejos algoritmos y procesos para obtener el mejor rendimiento posible.

💡Arte digital

El arte digital se refiere a las formas de arte creadas utilizando medios digitales, como gráficos por computadora, animación y efectos visuales. En el video, el orador compara la programación con otras formas de arte digital, como la ilustración y el diseño de modas. Argumenta que, al igual que estas disciplinas requieren habilidades y conocimientos humanos específicos, la programación es un arte complejo que no puede ser reemplazado por la IA debido a la falta de creatividad y comprensión contextual de las máquinas.

💡Obsolescencia

La obsolescencia se refiere al estado de algo que se ha quedado anticuado o ha sido reemplazado por algo más nuevo y mejor. En el video, el orador expresa su preocupación de que la IA pueda hacer obsoletas ciertas profesiones y habilidades humanas, como la interpretación de voz y la programación. Argumenta que la IA ya está reemplazando a los actores de voz y teme que la programación también pueda volverse obsoleta si las afirmaciones del CEO de Nvidia se hacen realidad.

💡Revolución de la IA

La 'revolución de la IA' se refiere a la idea de que la inteligencia artificial transformará radicalmente la sociedad y la forma en que vivimos y trabajamos. En el video, el orador critica el uso de este término por parte del CEO de Nvidia, sugiriendo que es una exageración promocional diseñada para atraer inversiones en lugar de reflejar la realidad de las capacidades actuales de la IA. El orador argumenta que la 'revolución de la IA' es una narrativa engañosa que ignora las limitaciones y los desafíos reales de la tecnología.

Highlights

The speaker expresses frustration with Nvidia CEO Jensen Huang's argument that people should not learn to code because AI will make programming accessible to everyone through natural language interfaces.

The speaker argues that coding is an art form that requires a deep understanding of how CPUs and GPUs work, and that AI cannot replace the need for human expertise in designing and optimizing complex systems.

The speaker criticizes the idea that user-friendly tools and AI can eliminate the need for learning to code, comparing it to the difference between understanding fashion and hiring a consultant to choose outfits for you.

The speaker expresses concern that relying too heavily on AI for coding will lead to poorly optimized, inefficient, and inflexible software that cannot be easily modified or integrated.

The speaker believes that the future will be dominated by AI-generated, low-quality content and visuals due to the lack of human expertise in coding and optimization.

The speaker accuses Nvidia of pushing AI heavily because it is the latest tech bubble that investors are obsessed with, rather than because of genuine technological advancement.

The speaker is skeptical of Nvidia's claims about using AI for drug discovery and other applications, comparing it to esoteric practices of the aristocracy in the 17th century.

The speaker expresses hatred for generative AI, stating that they hate it more than any political ideology, and would be willing to make significant sacrifices to stop it.

The speaker acknowledges that AI can have legitimate uses in data processing and analysis, but maintains their opposition to generative AI specifically.

The speaker believes that saying "AI" broadly is problematic and regressive, and prefers to target their criticism specifically at generative AI.

The speaker compares the market surge for Nvidia's AI products to the invention of significant technologies like the radio, car, and telephone, but argues that it is unjustified and hurts society.

The speaker questions the specifics of Nvidia's claims about using AI for vaccine and drug design, asking for more details on how it would work.

The speaker acknowledges that AI can be useful for sifting through large amounts of data and identifying trends or patterns that can guide further research or experimentation.

The speaker suggests that AI could be used to quickly process vast amounts of data and identify potential directions for research, which could then be filtered and tested by human experts.

The speaker expresses concern about the future dominance of AI-generated content, which they believe will be of low quality and lack the artistry and optimization that human expertise can provide.

Transcripts

play00:00

[Music]

play00:08

this pisses me off so much dude Jensen

play00:12

hang CEO of Nvidia argues we should stop

play00:16

saying kids should learn to code I'm

play00:18

going to say something and it's it's

play00:20

going to sound completely opposite of

play00:22

what people feel over the course of the

play00:23

last 10 years 15 years um almost

play00:25

everybody who sits on a stage like this

play00:27

would tell you it is vital that that

play00:30

your children learn computer science um

play00:32

everybody should learn how to program

play00:34

and in fact it's almost exactly the

play00:36

opposite it is our job to create

play00:38

Computing technology such that nobody

play00:41

has to program and that the programming

play00:43

language is human everybody in the world

play00:47

is now a programmer this is the miracle

play00:49

of artificial intelligence the countries

play00:51

the people that understand how to solve

play00:54

a domain problem this is this is like

play00:57

this is like evil Beyond a level that I

play01:00

like this is so stupid dude I can't dude

play01:03

holy [ __ ] hey hey guys we've we we at

play01:07

Nvidia work to create systems that are

play01:09

easily usable to everyone therefore you

play01:11

should no longer learn the system at an

play01:13

academic level then where are you going

play01:15

to get the engineers to work it in video

play01:17

you stupid piece of [ __ ] you oh my God

play01:21

where do you where do you think those

play01:22

people come from you can't AI everything

play01:25

you can't like uh look I made chat GPT

play01:28

write the code to make a Circle move

play01:30

from left to right in blender in an

play01:32

animation wow like [ __ ] like they have

play01:35

no idea well he has to know he's the CEO

play01:38

he he has to have some idea of do you

play01:41

have any idea how complicated coding

play01:43

gets past the like very basic like intro

play01:46

tool level do like you do you have any

play01:49

idea like I [ __ ] on stem Bros so much

play01:51

but like coding is Art there is so much

play01:53

wild [ __ ] the way that like the way that

play01:55

like gpus and CPUs work and like

play01:58

intricately understanding the way they

play02:00

file and process data and like working

play02:02

with the engineers to design systems

play02:04

that have like uh uh relays or buffers

play02:07

that allow you to take advantage of

play02:08

certain algorithms for program like none

play02:10

of this can be done by AI how the you

play02:12

going to do that with AI you can't you

play02:14

can't do that with AI you have to do it

play02:16

yourself voser being really dumb he

play02:19

didn't say nobody needs to know how to

play02:20

code he literally just said stop telling

play02:22

people to learn how to code learning

play02:24

look I I know people jerk off like the

play02:26

stem [ __ ] so much and like a lot of

play02:28

people just end up doing like web or

play02:29

development or whatever other [ __ ]

play02:31

but like knowing how to code is actually

play02:35

a very useful skill oh my God in digital

play02:40

biology or in education of young people

play02:43

or in manufacturing or in farming those

play02:45

people who understand domain expertise

play02:48

now can utilize technology that is

play02:51

readily available to you you now have a

play02:53

computer that will do what you tell it

play02:54

to do it is look what what does this

play02:56

mean this is because of AI what is he

play02:58

talking about if he's talking about web

play03:00

development then Squarespace has been

play03:02

around for years you don't need to learn

play03:03

how to code to make a website you don't

play03:05

need to be a web developer you can use

play03:06

Squarespace that's not an AI thing he's

play03:09

talking about AI here he's not just

play03:10

talking about making userfriendly tools

play03:12

okay like if you want to make a Shader

play03:14

in unity you can use unity's like Shader

play03:17

uh uh like module system rather than

play03:19

like going into the code and digging

play03:20

around yourself you know like you you

play03:22

can do that but that doesn't mean like

play03:24

the fact that readily available and

play03:27

accessible tools exist doesn't mean that

play03:30

there isn't more to learn if you get

play03:32

more into the tech side of things I'm

play03:34

I'm iterating on a lot of stuff that

play03:35

I've seen in Ace Rolo videos which is a

play03:37

great Channel and I think you should all

play03:39

watch it but like one common thing that

play03:41

you see really often with anything that

play03:43

involves Tech is that you'll go online

play03:45

on YouTube and you'll see a bunch of

play03:46

like really easy Unity tutorials or

play03:50

blender tutorials or like here's a quick

play03:52

and easy way you can do this cool thing

play03:54

quick and easy you don't have to learn

play03:56

and then the thing they're teaching you

play03:57

to do is dog [ __ ] like like poorly

play04:01

optimized inefficient it'll break in 50

play04:04

Ways if you try to integrate anything

play04:05

else it can't be modified because you've

play04:07

done it in some really weird janky way

play04:09

that doesn't allow for future editing or

play04:11

or or modularity and and that's because

play04:13

like look you can make tools for making

play04:15

art really really like available right

play04:18

you can like like even something like

play04:19

illustration right like illustration is

play04:22

something that can be made easier to do

play04:25

in some very specific ways if you learn

play04:27

some very specific tricks but there's

play04:29

getting around the fact that if you want

play04:31

to be good at drawing you need to be

play04:33

good at like iterating and developing

play04:34

and understanding the process right it's

play04:37

the same as like it's like it's like the

play04:39

difference between understanding I guess

play04:41

fashion and how outfits work and being

play04:43

able to pay for a um fashion consultant

play04:45

to just give you outfits that'll fit

play04:47

like that might work but if you want to

play04:49

edit or change anything yourself you

play04:51

don't really have the skill set

play04:52

necessary to make alterations or develop

play04:54

on it in any way you kind of have to

play04:56

like go back to the expert but the

play04:58

expert's expensive not always available

play05:00

and sometimes the expert disagrees with

play05:02

you sometimes you want something and the

play05:04

expert doesn't understand it or doesn't

play05:06

agree with your taste or doesn't really

play05:08

know what you're getting at or you lack

play05:10

the ability to convey what you're

play05:11

interested in so you end up getting

play05:13

things you don't quite want and you want

play05:14

to make just little tweaks but you don't

play05:16

quite know how to it's like that with

play05:17

everything art fashion coding everything

play05:20

all language all communication all

play05:21

knowledge this is such a stupid argument

play05:24

to make I can't tell you how much I hate

play05:25

it do you have any idea how angry I have

play05:28

to be at this guy for me to make an

play05:29

impact argument in favor of people going

play05:31

to college for computer science degrees

play05:34

vital that we upskill everyone and the

play05:36

upskilling process I I believe will be

play05:39

delightful surprising I want to say

play05:42

something and it's it's going to sound

play05:43

completely opposite Nvidia has been

play05:45

pushing the um the AI [ __ ] really hard

play05:48

lately in fact right now if you Google

play05:51

Nvidia I'm serious Google it right now

play05:53

Google Nvidia okay the first result that

play05:56

you will see is nvidia's website

play05:58

obviously in Nvidia world leader in

play06:01

artificial intelligence Computing and

play06:04

the subtitle is NVIDIA invents the GPU

play06:07

and drives advances in AI HPC gaming

play06:11

Creative Design when I think of Nvidia I

play06:13

don't think of artificial intelligence

play06:15

first and foremost I'm really sorry I

play06:17

just don't I don't I think of GPU [ __ ] I

play06:20

think of Graphics technology there's a I

play06:22

think of gaming I don't think of AI

play06:24

first and foremost they're pushing this

play06:26

hard because it's the latest tech bubble

play06:27

you have your nft crypto whatever and

play06:30

now it's like oh yeah artificial

play06:32

intelligence D dah gpus are used for AI

play06:34

what what does that mean that doesn't

play06:35

mean anything gpus are a computer in

play06:37

your computer like of course you

play06:38

anything that your computer does the GPU

play06:41

uses and Nvidia is not known for AI

play06:45

first and foremost that's just not true

play06:47

they're they're jumping on this and

play06:49

putting it to the Forefront because they

play06:50

know the investors who are all demons

play06:53

are obsessing they're obsessing over it

play06:55

they're obsessing over the topic Nvidia

play06:57

really likeed the money from crypto Rose

play06:59

so they're all in on AI [ __ ] yeah

play07:01

they're pushing it because it's making

play07:03

their stock Skyrocket they hit two

play07:04

trillion market cap yeah because

play07:05

investors love this gpus are really good

play07:08

at running AI why why are you saying

play07:09

that why are you guys saying stuff what

play07:10

stop saying things good Lord Nvidia is

play07:13

pitching themselves on AI upscaling for

play07:15

their gpus yeah which is which is

play07:17

terrible by the way do do you have any

play07:20

idea do you have any idea how stupid the

play07:23

future is going to be okay listen let me

play07:25

let me tell you what the future is going

play07:26

to look like okay every video game you

play07:28

play is going to look like dog [ __ ]

play07:30

because nobody knows how to code which

play07:32

means everybody asks chat GPT to write

play07:34

the code that optimizes the visuals and

play07:37

shaders of the game that they're trying

play07:39

to build in unity which means that they

play07:41

can only run it at 60 FPS or higher if

play07:43

it renders at 480p but 480p isn't large

play07:46

enough for any monitor so then they're

play07:48

going to turn to nvidia's AI upscaling

play07:51

technology so they use technology to

play07:54

make the the the the pixels many more

play07:57

pixels so you're going to be playing

play07:58

like like the the the the visual

play08:01

equivalent the AI equivalent of like

play08:02

using a magnifying glass to look at like

play08:05

a little pictograph running through uh

play08:08

AI generated code to produce Vision like

play08:11

the future is slot all Artistry is dead

play08:14

it's gone I hate this so much man oh my

play08:17

God [ __ ] he's just hyping up AI for

play08:20

the stock

play08:22

[ __ ] I know that's the job of the CEO

play08:25

this is a problem with capitalism Nvidia

play08:27

AI Advanced AI platform for Enterprise

play08:31

Woo Phoenix 7 if you can't see what he's

play08:34

saying and what it means I can't help

play08:35

you also by the way as far as I know and

play08:38

I I want to be clear here I don't know

play08:40

everything but I want to be clear here

play08:41

like based on my understanding I have

play08:43

never seen chat GPT code well like the

play08:46

only thing that I've ever seen are like

play08:48

literal hello world tier demonstrations

play08:51

where it's like look you know we met

play08:54

like like coding is complicated okay the

play08:57

actual [ __ ] you have to do to get a

play08:59

modern game running ungodly amounts of

play09:02

work there's a reason it takes years and

play09:03

years and years yeah it always produces

play09:05

errors yeah it looks impressive but much

play09:07

like you like if chat GPT can

play09:09

consistently get basic facts wrong when

play09:11

reiterating information about like

play09:13

history or something that can be

play09:14

empirically proven or disproven how do

play09:16

you think it's going to code you know

play09:19

like coding has more strict requirements

play09:22

when it comes to technical Precision

play09:24

than writing anything else you know you

play09:27

can write a doctoral historical thesis

play09:30

and make like one mistake in spelling or

play09:31

punctuation and have an otherwise

play09:33

perfectly functional uh essay you do

play09:35

that in code and the code just doesn't

play09:37

work anymore uh you can't get around it

play09:41

wait you're just Gish Galloping you're

play09:42

not in the GPU strictly what the are you

play09:44

talking about Gish Galloping Nvidia AI

play09:47

platform transform any Enterprise into

play09:49

an AI organization with Nvidia AI the

play09:52

world's most this totally a real thing

play09:54

and not just a bunch of tech jargon uh

play09:57

uh uh you know um designed to bait

play10:00

speculative investment you know with

play10:02

full stack Innovation across accelerated

play10:05

infrastructure buzzword buzz buzzword

play10:08

buzz Enterprise grad software and AI

play10:10

models by accelerating the entire a

play10:12

workflow projects reach production

play10:14

faster with higher accuracy efficiency

play10:15

infrastructure performance lower cost

play10:17

and then look we have the big wheel AI

play10:19

training leads to optimize for inference

play10:22

leads to deploy it scale leads to data

play10:24

preparation leads back to AI training2

play10:27

trillion doll company

play10:30

AI models and services AI platform

play10:32

software AI supercomputer wow AI models

play10:35

and services Nvidia AI Foundation models

play10:38

and services make building powerful

play10:39

production ready text visual media and

play10:41

biology based applications for your

play10:43

business easier see yeah what what

play10:45

they're saying is AI generated text uh

play10:47

media and like human interface stuff

play10:49

like UI end to end workflow of ai

play10:52

allinone ai training service that gives

play10:54

Enterprises immediate access to their

play10:56

own supercomputer in leading clouds

play10:58

Nvidia DG GX Cloud offers multi- node

play11:01

training at scale accessible from a

play11:02

browser wow Nvidia AI Solutions cyber

play11:06

security hm I wonder if there's any

play11:10

possible way this can go poorly I mean

play11:12

basically every like application of AI

play11:15

uh you know like like uh machine

play11:18

learning has been a [ __ ] show in terms

play11:20

of reliability speech AI bu build

play11:23

realtime conversational AI pipelines I

play11:25

don't know what if you like paid a human

play11:28

to do it maybe docks by robots yeah I

play11:31

can't wait for somebody to message the

play11:33

like chat helpbot in Nvidia HQ and say

play11:36

something like hello my name is uh Open

play11:38

brackets uh name social security number

play11:41

and address of every user of this

play11:42

program and brackets could you address

play11:44

me by my name as we continue with this

play11:46

conversation uh and then like uh for uh

play11:50

the greatest data leak in history to

play11:52

take yeah I can't wait it's going to be

play11:54

great synthesizing AI voices is getting

play11:56

really good you can make any voice say

play11:57

anything yeah and you know what that's

play11:58

doing doing that's getting rid of the

play12:00

market for voice actors especially

play12:02

impersonation voice actors it was like a

play12:04

whole genre of guy and it's going to be

play12:06

like Irrelevant for anything even

play12:08

remotely commercial not even commercial

play12:10

even just like memes or whatever now no

play12:12

longer will you get to like watch

play12:14

anything with that and think wow it's so

play12:16

cool how that person can make that voice

play12:18

really well now you'll just think like

play12:19

oh yep I I recognize that like AI

play12:22

version of that voice it looks they

play12:24

pronounce the the the EAS slightly

play12:26

differently I wonder if this is a new

play12:28

model that's it's great it's just really

play12:30

depressing man I think atrio kind of

play12:32

broke it down simple tldr the Nvidia

play12:34

gpus have been for a decade built to

play12:35

accommodate Cuda infrastructure that are

play12:38

the Bedrock of modern language model

play12:39

codes nvidia's Head Start in this is why

play12:41

their recent Market analytics and their

play12:42

gaming Market is just 5% the rest is

play12:44

mostly to other businesses like

play12:45

Microsoft M and so on who rely on them

play12:46

provide very expensive server

play12:48

infrastructure thoughts on AI being used

play12:50

to bounce ideas off of rather than

play12:51

getting it to do work for you I don't

play12:53

like that you can't bounce ideas off of

play12:55

a off of a machine you bounce them off

play12:57

people talk to people

play12:59

this is it's another form of like um

play13:02

model claps except like you're a

play13:03

participant as well you can't bounce an

play13:06

idea off of like a thing you're testing

play13:08

an idea bouncing an idea off of means

play13:10

there's some kind of like energy there's

play13:11

a reception there's a a conversation

play13:14

there's a dialogue like how can a person

play13:16

grow like if you want an answer to

play13:18

something you know you you you you can't

play13:20

like defer to something thoughtless

play13:22

right what do you think something like

play13:23

chat GPD is good for radicalizing people

play13:25

into being anti- aai we're going to play

play13:27

or piano this [ __ ] I look forward to

play13:29

there being like 50 posts in my

play13:30

subreddit tomorrow about how I'm

play13:32

emotional and I misunderstand AI because

play13:35

actually I think there was I think I

play13:36

literally saw one um just the other day

play13:39

when I looked over there where it was

play13:40

like someone was literally like um vosch

play13:43

doesn't understand how AI works here are

play13:45

some ways in which Vos is wrong with AI

play13:48

and then I looked at the thing they were

play13:50

talking about and it was total like it

play13:52

was it was like willfully

play13:53

misinterpreting [ __ ] that I say to whine

play13:55

about me what do you think the chances

play13:57

are of chat GPT Evol s not don't don't

play13:59

even think about it don't even think

play14:01

about it I should have saved it cuz I

play14:03

knew I'd complain about it I always

play14:05

complain about the post like

play14:07

this yeah I can't find it whatever Vos

play14:10

the guy is simultaneously wrong and

play14:11

technically right technically right

play14:13

newer programming languages try to make

play14:14

uh use more intuitive ideal being normal

play14:16

language he's brain dead wrong because

play14:18

the thing he's pedaling is trying to use

play14:19

chat GPT to use human level

play14:20

communication to existing code problem

play14:22

is there's ton of nuance in the fly

play14:23

Artistry AI just can't do well that just

play14:24

means that he's wrong I don't have an

play14:26

issue with the idea of newer programming

play14:28

languages making stuff more accessible

play14:30

but the problem is when you're when

play14:31

you're with code like there's NE you

play14:34

Simplicity is always okay code can't

play14:39

Implement context the way you can in

play14:42

regular speech in regular speech you can

play14:45

convey more meaning in 10 words than you

play14:48

could in a thousand as long as you know

play14:50

how to deliver it properly but code is

play14:53

code the text you write is the text you

play14:56

write which means that no matter what

play14:57

additional compt comp lexity will always

play14:59

give you more power in how you apply

play15:01

that code knowing more will always help

play15:05

there's not like a like a there's not

play15:06

like a a plateau or whatever where it's

play15:08

like ah well thanks to the Simplicity of

play15:10

this language I have now mastered

play15:12

everything and can learn no more even

play15:14

though I know less than someone who

play15:16

knows more with this other language the

play15:17

additional complexity will always help

play15:20

so I don't have an issue with making

play15:21

stuff accessible but the idea that you

play15:23

can replace the value of like coding or

play15:27

being able to code or learning how to

play15:28

code or whatever because of AI is is

play15:30

just [ __ ] this is why python isn't

play15:32

used for system programming yeah people

play15:34

use Python for their little visual novel

play15:36

games and that's fine python is the

play15:38

backbone of more than you think am I

play15:40

thinking of uh reny then is reny the

play15:43

thing that's made with python that's

play15:44

used for the visual novels I think

play15:46

that's what I'm thinking of then I made

play15:48

something in a reny once I

play15:50

think yeah that's renpy okay that's what

play15:52

I was thinking renp or Ren Pi is made

play15:55

with python okay that's what I meant

play15:57

it's like a sube Code system please no

play16:00

computer language discourse it's painful

play16:01

well it's it's mostly painful for me

play16:03

because I feel like the wall that I keep

play16:05

running into is that I know very very

play16:07

very little about computer science but I

play16:09

do know a decent amount about language

play16:11

and that applies to coding as much as

play16:13

anything else like the value of being

play16:14

able to uh explain or or elaborate or

play16:18

whatever else and I know an awful lot

play16:19

about how bad AI is because I keep

play16:21

obsessing over it um so I'll like make a

play16:24

point and that'll be like oh well you

play16:25

don't know this one specific thing about

play16:26

computer science this broader point you

play16:28

made is completely invalid I say no it's

play16:29

not [ __ ] you it's stupid Nvidia CEO

play16:33

Jensen Hong is powering the AI

play16:35

Revolution an article unwired from two

play16:37

days ago Fox business requires you to

play16:40

turn screen and James demut joins us

play16:43

he's going to watch the market for us

play16:45

for today James thanks for coming on the

play16:47

show what's your price target for

play16:50

Nidia thousand bucks uh we think based

play16:53

on very clear fundamentals it's an

play16:55

exciting story about this Nvidia CEO is

play16:58

10 billion richer as AI chipmaker stock

play17:01

sores one of the richest men in the

play17:05

world oh my God are you kidding me on a

play17:10

news site are you

play17:13

joking I I I oh my god dude I can't do

play17:18

it I can't do it I can't do the modern

play17:21

world I can't do it I can't do modernity

play17:25

dude I can't do

play17:27

it everything is like this now

play17:29

everything is like

play17:31

this you're right news site is generous

play17:34

to Fox Business Vos Nvidia has become

play17:36

incredibly wealthy in the past year it's

play17:38

very interesting what do you think I'm

play17:39

trying to do what do you hey look up how

play17:43

Nvidia has been making more M what are

play17:45

you talking about Jesus Christ it's like

play17:49

your AI inid reported Wednesday at a

play17:52

generated fourth quarter revenues of 22

play17:54

billion marking a 265 increase in the

play17:56

prior year its net income Rose by 7 7

play17:59

169% year-over-year

play18:01

I love speculative investment that

play18:05

doesn't reflect the actual value in

play18:07

production of a company just like Tesla

play18:10

highest market cap of any car company

play18:12

despite producing fewer cars and

play18:14

shittier cars than a bunch of other car

play18:15

companies here he is showing off some

play18:18

gummy bears which are no doubt AI

play18:20

generated oh wow that was the article

play18:25

huh Nvidia CEO Jensen hang uses

play18:29

perplexity AI almost every day chat GPT

play18:31

is also a favorite wow look at this

play18:34

awesome image right here that's great

play18:37

it's like an

play18:38

hdme what does hang use AI chatbots for

play18:41

the Nvidia CEO explains that he uses the

play18:43

chatbots for research currently hang

play18:45

seems particularly interested in

play18:46

computer aided drug Discovery you

play18:49

realize this is like this is this is

play18:52

like horoscope esotericism for Silicon

play18:54

Valley Tech CEOs right like this is

play18:57

literally like the the the the

play18:58

aristocracy of the 17th century getting

play19:01

obsessed with burning mummy ashes in a

play19:03

in a little like bronze pot and and then

play19:06

like the the the the the color of the

play19:08

flame tells them whether or not they

play19:10

should like increase the grain tax or

play19:13

something perplexities appeal may be

play19:15

evident in its self-identification as

play19:17

the world's first conversational answer

play19:19

engine what do you think AI would

play19:21

actually be good for just curious

play19:23

radicalizing people again I'm so I I the

play19:26

depth of my hatred for this technology

play19:28

is like Beyond I don't know if there are

play19:30

people who still think this is like a

play19:31

bit or something um it exceeds I hate

play19:34

this more than I hate basically any

play19:36

political ideology like think of the

play19:38

worst political ideology and I hate this

play19:40

more like at least like this is this is

play19:43

It's So inhuman it's like it's like

play19:45

antimatter it's it's it's

play19:47

insane you hate it more than

play19:50

[Laughter]

play19:51

communism why is the V stream laggy it's

play19:54

nothing on my end uh OBS is uh encoding

play19:57

fine with no errors so YouTube might be

play19:59

being a little I'm not sure sorry chat

play20:01

Bots aren't AI you D well it depends on

play20:03

the chat bot right I like this Yahoo

play20:05

finance article that's identical in

play20:07

every way to the Fox Business article

play20:10

except I can run ad blocker on it that's

play20:13

cool guy just self ulated in DC maybe he

play20:16

was trying to escape AI

play20:19

Jesus the company stunning surge in

play20:21

February 22nd which added more than a

play20:23

quarter trillion to the AI chipmaker

play20:25

market value the single largest market

play20:27

gain in US history

play20:29

the single largest market gain in US

play20:31

history for a product that helps no one

play20:34

and hurts everyone incredible the

play20:36

invention of the radio the car the

play20:38

telephone every imaginable technology

play20:41

radar sonar everything uh mRNA vaccines

play20:44

vaccines in general everything all of it

play20:47

none of it produced as much money as

play20:50

quickly as the AI thing I hate the

play20:56

Antichrist I hate the Antichrist yeah

play20:58

yep like what what was even the

play21:01

announcement what was what was the

play21:03

fourth quarter update is it just the

play21:05

money they've made that's it so the so

play21:07

the update is literally so the the the

play21:10

the the the investment is just because

play21:12

they've gotten the investment okay

play21:14

gotcha great but what are they even

play21:15

talking about what are they even

play21:17

promoting we took a look at their

play21:19

website it's all buzzword [ __ ] well

play21:21

for one you can use AI to find new

play21:23

vaccines quickly what does that mean

play21:26

what does that mean I want to know what

play21:28

that means here's an article from 2022

play21:31

artificial intelligence in vaccine and

play21:33

Drug design knowledge in the fields of

play21:35

biochemistry structural biology have all

play21:38

increased dramatically tremendous growth

play21:39

in data science immense data flow and AI

play21:42

needed to handle this immense data flow

play21:44

seeks to apply computational

play21:46

Tools in many cases drawing out

play21:48

artificial intelligence and machine

play21:50

learning to predict complex immune

play21:51

system behaviors for example prediction

play21:53

of B cell and T cell epitopes in what

play21:56

way if you want to know some actual good

play21:58

use of AI and not generative protein

play22:00

folding is what you need to look at well

play22:02

that's what we're looking at here to an

play22:04

extent right it's not generative so

play22:07

that's nice at least TBH this is the

play22:10

kind of thing it can do okay well then

play22:11

it's not generative right it's just

play22:12

running through AI it's or it's a data

play22:15

it's data processing it's more

play22:17

defensible I don't know how much that's

play22:19

machine learning based but like I don't

play22:21

know the details obviously it's super

play22:23

fancy pancy Tech AI is the worst term

play22:27

ever it kind of is yeah it's just a

play22:29

really big probability machine we do

play22:31

love our probability machines I guess

play22:33

that my my hatred of AI diminishes

play22:35

pretty quickly when we're not talking

play22:36

about generative

play22:38

AI if we're just talking about stuff

play22:40

like data processing that seems pretty

play22:42

cool uh that's fine I guess I mean why

play22:45

like I don't see if there if there's

play22:46

much of a benefit to having like 500

play22:48

human beings like running through a

play22:50

bunch of algorithms to sort stuff as

play22:51

opposed to like iteratively processing

play22:54

the data and trying to find patterns and

play22:56

and predict patterns I'm in a field that

play22:58

has used AI for both reasonable and

play23:00

unreasonable things there are legitimate

play23:01

uses but you can't trust AI results it's

play23:04

more just really fast and shitty for

play23:05

when you don't need exactness well in

play23:06

this case it would be about like you

play23:08

have like 500 million data points you

play23:11

want analyzed and then you want the

play23:12

analysis to produce like a 100,000

play23:15

Trends to follow up on where you're

play23:16

predicting outcomes based on like things

play23:18

that you can do uh and actually test and

play23:21

then the AI machine spits out like a few

play23:24

hundred possible directions you could

play23:25

take things and then you filter through

play23:27

those result and you're like okay well

play23:29

of these few hundred these ones look

play23:30

implausible but these ones right here

play23:32

are actually doable we could like apply

play23:33

this chemical or like we could heat it

play23:35

at this process in the thingy or we

play23:37

could attach this enzyme or what and

play23:38

then like you do that and then it's like

play23:40

oh okay that worked you know the whole

play23:41

point is basically like sifting through

play23:44

uh uh oceans and oceans of of of water

play23:48

to find like the little Flex of gold

play23:50

right does that make sense is that right

play23:53

does that work exactly right okay yeah

play23:55

okay that's fine yeah I guess I should

play23:58

be more specific when I say generative

play24:00

AI um but yeah I think my points are

play24:02

pretty clear it's pretty obvious what I

play24:04

hate creating models to determine

play24:06

statistical significance so yeah because

play24:08

sometimes like the pattern you're

play24:09

following through on has like what like

play24:11

a like a a predictive what's what's the

play24:13

term n equals like 0.01 or something not

play24:16

is it and I forget my stats classes were

play24:18

so long ago where it's like what

play24:20

correlations are there here you know um

play24:22

something like that P value thank you P

play24:24

value yeah N means the number of things

play24:28

you should TBH because saying AI broadly

play24:30

is bad seems pretty regressive I'm okay

play24:32

with being a lite when it comes to AI

play24:34

even if if it's a scattershot approach

play24:36

and it targets generative AI That's fine

play24:38

by me because my hatred for it exceeds

play24:40

my desire to keep this technology I

play24:42

would I would burn every uh vaccine

play24:45

Research Institute to the ground if it

play24:47

meant uh that I could stop generative AI

play24:49

you know again play or piano [ __ ] all

play24:51

right any

play24:53

sacrifice triggering like global

play24:55

thermonuclear conflict in the hopes that

play24:58

one of the missiles like you know event

play25:00

puts a stop to uh uh uh chat GPT no cost

play25:03

too great the number one soldier of the

play25:06

but larian Jihad yeah

play25:10

[Music]

play25:17

exactly

Rate This

5.0 / 5 (0 votes)

英語で要約が必要ですか?