Shadowing Practice: AI Destroyed My Profession - Learn English Speaking with Video

Elige un modo de práctica

Cada modo abre una sala de práctica a pantalla completa.

195 oraciones

Si las oraciones son demasiado cortas o largas, haz clic en Editar para ajustarlas.

It's December 2013 and the President of the United States is looking straight into a camera talking to kids all over America.

He tells them to stop playing video games and instead to start making them.

And then he says a sentence

that ends up on classroom posters all over the country

and you know what a whole ass generation listened coding boot

camps popped up everywhere computer science became one of the most popular majors

and the phrase learn to code turned into the safest career advice on the planet it got

so popular that in 2019

when a bunch of journalists lost their jobs people online mocked them with just three words learn to code

now fast forward to 2026 in the first half of this

year alone tech companies announced almost 140 000 job cuts

and most of them say the reason is ai so is that it is the best job of the 21st century dead

Let me keep it real with you folk, I'm asking that question sincerely,

I'm a software engineer myself with a master's degree and everything so for me this isn't just another story from the news,

it's my industry and honestly that makes this video a little personal.

Here's the thing, companies thought that AI would be a cheap replacement for programmers

And now that they've actually tried it, three bills are starting to show up.

One for the AI itself, one for the code it writes, and a third one that almost nobody is talking about yet.

So let's open them up, one by one.

Okay, let's start with the layoffs, because they're very much real.

In January, Amazon cut 16 ,000 corporate jobs.

Oracle let go of around 21 ,000 people over the year,

Meta cut 8 ,000, and a payments company called Block laid off half of its entire workforce in one go.

And the people building AI didn't really help calm anyone down.

Back in 2025, the head of Anthropic, one of the biggest AI companies in the world said AI

could wipe out half of all entry -level office jobs within a few years.

So it's easy to look at all of this and think that it's over, that we lost.

But not everyone believes the companies are telling the whole truth.

Even Sam Altman, the guy who runs OpenAI, said that some companies are AI washing their layoffs.

In other words, They blame AI for cuts they were going to make anyway.

And you have to admit, it's a pretty convenient excuse, because saying we fired 4 ,000 people just sounds terrible.

But saying that AI changed our industry sounds like the future.

AI is the perfect excuse, because it makes the decisions look like it came from the technology and not from the company.

There's also one more thing a lot of headlines leave out.

Many of these companies are spending insane amounts of money on data centers and AI chips right now.

So sometimes when a company says AI took your job,

what it really means is we needed your salary to pay for AI.

But is programming actually dying?

Before we open any of those bills, let's check that first, because the answer is not exactly what the headlines say.

In 2025, one headline went all over the internet.

The number of programmers in the US had dropped to its lowest level since 1980.

That's a time when most people didn't even know what the internet was.

The numbers are correct, but there's a catch.

You see, in the official US statistics, a computer programmer is a pretty small category.

It's around 110 ,000 people whose job is mostly writing code based on someone else's plans.

Most of the people you'd call programmers are actually counted as software developers and that group has around 1 .9 million people.

The official forecast says it will grow by 10 % over the next 10 years.

So the headline wasn't exactly lying, it just compared one small room to the whole building.

It's the same story with the chart saying computer science graduates have one of the highest unemployment rates out of any major.

The sample behind it was very small, and when economists took a closer look,

it turned out that around 90 % of young people with that degree were actually working.

So no, programming is not dead.

But something is definitely wrong.

And you can see it the moment you follow the money.

Time to open the first envelope.

The bill for the AI itself.

First, a quick explanation.

AI tools charge by something called tokens.

A token is basically a small piece of a word

and you pay for every token the AI reads and every token it writes.

So, the more work you give it, the more it costs.

Now, here's the funny part.

Over the last few years the price per token has dropped like a rock

and the bills they actually went up how is

that possible AI coding tools don't just finish your line of code anymore

now they work on their own code for hours reading through huge amounts of code

and trying things again and again so even though every single token is cheap, they burn through an unbelievable amount of them.

It's like gas getting 10 times cheaper and then everyone going out and buying a private yacht.

And in 2026 companies started getting the bills.

In May, Uber's CEO admitted

that the entire company had burned through its entire yearly budget for AI coding tools in just four months.

He even said that the link between all that spending and any real benefit for a customer is not there yet.

Soon after, Uber put limits on how much each employee could spend.

Then in August it was Microsoft's turn.

It came out that one engineer had spent 28 ,000 dollars on AI in just 28 days.

That's a thousand dollars a day.

At that pace, one engineer's AI use would cost more throughout the year than paying three American software developers.

So Microsoft sent out an internal email with a sentence

that's honestly peak email said token maxing is not what we are optimizing for token maxing yep

that's an actual word in a real email in a trillion dollar company

and when your company has to put a word like

that in an official email just to tell you to stop spending money you know something went wrong

and it's not just uber

and microsoft gartner one of the biggest research firms in tech predicts

that by 2028 ai coding costs will be higher than the average developer's salary think about

that for a second the whole point was to replace expensive programmers with cheap AI,

and now the AI is on track to cost more than the programmer.

But hey, at least all that money bought a lot of code, right?

Well, that's exactly where the second bill comes in.

Envelope number two, the code.

To understand this one, you need to know a term programmers use all the time and that's technical debt.

It works exactly like a credit card.

When you write code very quickly it gets a bit messy.

You get something that works today and it feels amazing just like swiping a credit card.

But every change you make after that gets slower and more expensive because you keep paying interest on that mess.

And eventually, someone has to pay off the whole thing.

So what happens to that debt, when AI writes a big part of the code?

A company called GitClear has been tracking exactly that.

They looked at more than 600 million code changes from 2023 to 2026, and they found two things moving in the opposite directions.

100

Copying and pasting code went way up.

101

Meanwhile, the work of cleaning up and reorganizing old code,

102

which programmers call refactoring has collapsed in 2022

103

it made about 21 % of all changed code by 2026

104

it was under 4 % in other words AI is really good at adding new code

105

and really bad at tidying up the code that's already there

106

so the mess keeps growing and then there's security A company called Veracode tested more than 150 AI models,

107

and when nobody specifically told them to be careful,

108

about 45 % of the code they wrote had a known security hole hiding in there.

109

And you know what's the worst part?

110

That number has barely moved in two years.

111

The models got much better at writing code that works, but they didn't improve much at writing code that's safe.

112

Another big study looked at more than 300 ,000 changes made by AI tools on GitHub.

113

Almost a quarter of the problems the AI introduced were still sitting there in the latest version of the code.

114

Nobody had gone back to fix them.

115

AI really does make a lot of programmers faster, and that's a big deal.

116

And nobody has a reliable estimation of how much all of this debt will cost in dollars.

117

The bill simply hasn't arrived yet.

118

So yeah, AI is writing code faster than ever, but a lot of it is the kind of code

119

that someone some day has to come back and fix

120

which raises another question who's gonna fix it because that's the third bill

121

and it's the one no one is ready to pay

122

welcome to the third envelope and this one isn't about money or code it's about people.

123

In 2025 a team from Stanford started tracking payroll data for millions of American workers and they've kept updating it ever since.

124

What they found is

125

that young people aged 22 to 25 in jobs where AI can do a lot of the work

126

now have about 19 % fewer jobs than they should have compared to young people in other jobs.

127

But get this, experienced worker in the exact same jobs are doing fine.

128

Companies are not firing their seniors, they've actually quietly stopped hiring juniors.

129

And if you think about it, That makes sense.

130

The tasks companies used to give to juniors, for example, building simple features, writing tests, and fixing small bugs,

131

are exactly the tasks that the AI is best used for.

132

Researchers in South Korea talked to developers about this, and one senior engineer said something pretty chilling.

133

Their company let a lot of junior engineers go,

134

and there was, quote, no impact whatsoever.

135

Then the researchers talked to juniors who are still learning, and one of them said this.

136

I don't know what not to do, I've only seen the good examples.

137

See that right there?

138

That's the problem.

139

Juniors used to learn by writing bad code, breaking things and then getting roasted in code reviews.

140

Now the AI hands them the code that works and they never find out why it works.

141

Anthropic, one of the companies that actually makes these AI tools, tested this in early 2026.

142

They had 52, mostly junior developers, learn a new coding tool.

143

Half of them could use AI and the other half had to figure it out on their own.

144

Then everyone took a test to see how much they actually understood.

145

The group with AI scored 17 points lower and they weren't even noticeably faster.

146

Now think about where this goes.

147

Every senior developer on the planet started as a junior, who at first broke something, he made a mistake.

148

That's how you learn and right now we're building a system where nobody gets to break anything.

149

It's like a restaurant where the head chef gets an amazing machine that does all the chopping,

150

so the restaurant stops hiring kitchen staff.

151

It works great, right up until the day the head chef retires

152

and nobody in the building knows how to cook

153

so in 10 years who's gonna be the senior developer who

154

checks what the AI wrote who's gonna find those security holes

155

and clean up all the copied code from the bill number two put it all together

156

and you get something like this the AI writes the code then the code needs fixing

157

but the people who could fix it are never getting hired.

158

You can stop thinking of this third thing as a bill, instead think of it as an actual time bomb.

159

You know, there are people at the very top that can see it coming.

160

Amazon cut 16 ,000 corporate jobs in January 2026.

161

But that same year, the boss of Amazon's cloud business said that Amazon is hiring 11 ,000 interns and new graduates.

162

And when people asked him about replacing juniors with AI, he called it one of the dumbest things I've ever heard.

163

Say what you want about Amazon, but here, that's a good call.

164

So, what does all of this mean for you?

165

If you're a student, or you're thinking about learning to code, that same antropic study has some good news.

166

Not everyone in the AI group did badly.

167

The people who asked AI why something works, instead of just asking to do all the work for them

168

learned almost as well as the people who didn't use AI at all.

169

AI isn't your enemy, it's just a really bad teacher if you only ask it for answers.

170

Instead you should question everything and the most important question to ask is why?

171

Why is it making those specific decisions?

172

Use it for brainstorming and understanding

173

And suddenly, it might become one of the best teachers you'll ever have

174

The job itself is changing also

175

The time of manually typing out all of your code line by line is over You shouldn't do that

176

Unless you want to look like a Neanderthal to your own colleagues

177

Instead, you should focus a lot more on understanding what needs to be built, and checking whether the AI actually built it correctly,

178

which, if you've been paying attention, is exactly the skill that companies are running out of.

179

I can't promise anyone a job, and honestly, nobody can right now.

180

but if the data is right the companies

181

that fired everyone to save money are gonna need people who really understand code

182

and there's a real chance those people will be very very expensive okay remember

183

that sentence from 2013 back then it was the best career advice in the world

184

In 2026 it might need a small update.

185

Don't just let AI write your code.

186

Understand it.

187

Because programming is not dead.

188

But somebody is gonna have to pay these three bills.

189

And my guess is that it won't be the AI.

190

Unless of course the AI gets so good that it can clean up its own mess.

191

And to be fair, that's the one big question nobody can really answer just yet.

192

AI is getting better at coding pretty much every month.

193

Maybe one day it really will be able to do all of this on its own or maybe

194

today's AI is gonna hit a ceiling and will end up needing people who understand code more than ever.

195

Honestly, nobody knows, but that's a story for another video.

Descargar app

Todo lo que necesitas para hablar con fluidez

Pronunciación con IAEvalúa cada oración

Práctica de IPADomina cada sonido

VocabularioConstruye tu banco de palabras

Juego de vocabularioAprende mientras juegas

Escanea el QR

Download on theApp Store Get it onGoogle Play

FOR YOU

Recomendar Video

828 oraciones

Vocabulario y notas de pronunciación para esta lección

Esta lección de conversación se basa en el vídeo “AI Destroyed My Profession”. Las palabras que más se repiten: code, companies, bill. Este vídeo tiene 195 frases y 2476 palabras para practicar shadowing. La parte hablada dura 20:17. El hablante mantiene un ritmo constante de unas 122 palabras por minuto, cómodo para el shadowing. El 87 % de las palabras está entre las 3.000 más comunes del inglés, por lo que el vocabulario es fácil de seguir.

Vocabulario clave de este vídeo

15 palabras del vídeo que vale la pena aprender, con su pronunciación y significado:

Palabra	Pronunciación	Significado
programmer sustantivo	/ˈpɹoʊɡɹæmɚ/	programador, programadora
token sustantivo	/ˈtoʊkən/	seña
developer sustantivo	/dɪˈvɛləpɚ/	desarrollador
hire verbo	/ˈhaɪɹ/	contratar, emplear
headline sustantivo	/ˈhɛd.laɪn/	titular, encabezado
envelope sustantivo	/ˈɛnvə(ˌ)loʊp/	sobre, envelop
salary sustantivo	/ˈsæləɹi/	salario, sueldo
chef sustantivo	/ʃɛf/	chef
researcher sustantivo	/ˈɹiˌsɝ.t͡ʃɚ/	investigador, investigadora
meanwhile adverbio	/ˈmiːnwaɪl/	mientras tanto
chart sustantivo	/t͡ʃɑɹt/	chart
cloud sustantivo	/ˈklaʊ̯d/	nube, nimbo
wash verbo	/wɑʃ/	lavar
badly adverbio	/ˈbæd.li/	malamente, mal
explanation sustantivo	/ˌɛkspləˈneɪʃən/	explicación

Phrasal verbs que vas a escuchar

Palabra	Pronunciación	Significado
clean up verbo		limpiar
come out verbo	/ˌkʌm ˈaʊt/	revelarse, salir a la luz
end up verbo		terminar, ir a parar
find out verbo		averiguar, anoticiarse
go out verbo		salir
go up verbo		subir
lay off verbo		despedir, destituir
leave out verbo		dejar de lado

Gramática en este vídeo

Las estructuras que más usa el hablante, con las palabras exactas del vídeo:

Estructura	En el vídeo
Present perfect continuous have/has been + -ing — una acción que empezó antes y continúa	has been tracking · you've been paying
“Used to” used to + verbo — un hábito o estado pasado que ya no es cierto	used to give · used to learn
Present perfect have/has + participio pasado — una acción pasada que sigue importando ahora	they've actually tried · has dropped · has collapsed
Oraciones de relativo who / which + oración — información extra sobre una persona o cosa	guy who runs · people who could · people who asked

Pronunciación a tener en cuenta

El hablante usa 36 contracciones y formas reducidas, como gonna, didn't, don't. Dilas en su forma corta, tal como las oyes.

Los sonidos de “sh” y “zh”: chef /ʃɛf/, wash /wɑʃ/, explanation /ˌɛkspləˈneɪʃən/, estimation /ˌɛs.təˈmeɪ.ʃən/
Palabras largas — cuida el acento: developer /dɪˈvɛləpɚ/, explanation /ˌɛkspləˈneɪʃən/, unemployment /ʌn.ɪmˈplɔɪ.mənt/, unbelievable /ˌʌnbɪˈliːvəbl̩/, whatsoever /ˌ(h)wʌtsoʊˈɛvɚ/

Sonidos difíciles para hispanohablantes:

s + consonante al inicio — sin añadir una “e” delante: skill /skɪl/, swipe /swaɪp/
/v/ — no es /b/: los dientes tocan el labio inferior: developer /dɪˈvɛləpɚ/, envelope /ˈɛnvə(ˌ)loʊp/, convenient /kənˈviː.niənt/, unbelievable /ˌʌnbɪˈliːvəbl̩/, whatsoever /ˌ(h)wʌtsoʊˈɛvɚ/
/z/ — sonora, no /s/: phrase /ˈfɹeɪ̯z/, optimize /ˈɑptɪmaɪz/

Cómo practicar con este vídeo

Escucha el vídeo entero una vez sin hablar y anota las palabras que no conoces.
Haz shadowing frase por frase a velocidad normal, repitiendo cada una hasta que tu ritmo coincida con el del hablante.
Grábate y compara con el original, prestando atención a palabras como programmer, token, developer.

¿Qué es la Técnica de Shadowing?

Shadowing es una técnica de aprendizaje de idiomas respaldada por la ciencia, desarrollada originalmente para la formación de intérpretes profesionales y popularizada por el políglota Dr. Alexander Arguelles. El método es simple pero poderoso: escuchas audio en inglés nativo y lo repites en voz alta de inmediato, como una sombra que sigue al hablante con solo 1-2 segundos de retraso. A diferencia de la escucha pasiva o los ejercicios de gramática, el shadowing obliga a tu cerebro y músculos de la boca a procesar y reproducir simultáneamente patrones de habla reales. Las investigaciones muestran que mejora significativamente la precisión de la pronunciación, la entonación, el ritmo, el habla conectada, la comprensión auditiva y la fluidez al hablar, convirtiéndola en una de las metodologías más efectivas para la preparación del IELTS Speaking y la comunicación en inglés en el mundo real.

Técnica de shadowing: lee la guía completa paso a paso →

Shadowing Practice: AI Destroyed My Profession - Learn English Speaking with Video

Elige un modo de práctica

Descargar app

Harvard AI Professor: The People Most at Risk From AI Aren’t Who You Think

Did the AI bubble just pop?

It’s not just you

Why the Arab World never united

Virtual love - How dangerous are AI relationships? | DW Documentary

A Practical Guide To Becoming An AI Engineer (2026)

The Only SaaS Moats Left in 2026

The (Overdue) Collapse of AI Layoffs (here's the proof)

Jack Dorsey’s New Dessa AI Voice Sounds Scary Human

AI hype is starting to sound like religion | Tim O'Reilly

Naomi Klein: The AI Bubble Is Going To Burst. This Is How We Prepare.

You're Not Behind: Become AI-Native (without the overwhelm)