China’s AI Shockwave: A Wake-Up Call for the West?
Did we just experience a Sputnik Moment?
On January 10th, 2025, the internet was taken over by the story about the release of a new generative AI model from DeepSeek, a Chinese company nobody had ever heard before. The model was named DeepSeek-R1.
The news completely blindsided US tech companies, and the impact on their stock price was huge. NVIDIA alone lost 12% of their stock value in a single day.
This massive reaction wasn’t simply due to the announcement of a new model—new AI models pop up every week these days—what was shocking was the information that started to trickle out of what went into creating this model, specifically, the cost.
DeepSeek apparently achieved this complete AI model with less than one tenth of the price tags that any of the big players on the market (OpenAI, Meta, Google, etc.) had to spend. What makes it even more astonishing is that the USA actually currently has China under a selective sales embargo. This prohibits China, and any China-based company, from purchasing any latest-gen chips (as the USA is their main producer).
This means DeepSeek’s success was achieved without using any of the thousands of extremely expensive GPUs their American counterparts used for the training of all their own respective models. According to recent reports, they seem to have just used a significant number of older-generation GPUs.
Keeping all of these factors in mind, the total reported cost for training DeepSeek-R1 is a mere six million dollars.
Google is estimated to have spent around $191 million on training its Gemini Ultra AI model. Sam Altman stated that the cost of training GPT-4 was more than $100 million. Right now, we are still waiting for all these statements from DeepSeek to be confirmed and fact-checked, but, until this data is contradicted, this is their story. That is, of course, if it is contradicted.
To add insult to injury, DeepSeek was founded in 2023. As a company, it is a mere two years old. A startup that by all rights should be in its infancy was able to come out with a competitive product against the big tech companies who’ve been working on models like this for much longer and with exponentially larger capitals to draw from.
I’ve seen several people compare this huge upset and surprise within the American tech community to the 1957 announcement of the first satellite launch by the USSR Space Agency.
They posed the question: “Are we living through a new Sputnik moment?”.
So, is it really? Let’s break it down.
In my opinion, this should not have taken us by surprise at all.
If we had only paid proper attention to the massive advancements of Chinese industry in the last few years, we should have absolutely seen this coming.
Let me share just a few of them:
As of the end of 2023, China's high-speed rail (HSR) network was 45,000 km (27,960 mi) long. China also recently unveiled the prototype for the CR450 high speed train, so named because its max speed is 450 kilometers per hour (279 mph). The CR450 will operate commercially at 400 kilometers per hour (249 mph), making it the fastest train in the world.
In comparison, Amtrak's Acela is the only true high-speed rail service in the United States, reaching speeds of up to 160 mph (255 km/h) over 49.9 miles (80.3 km) of track.
The BYD Shenzhen is the largest cargo ship for automobiles in China. It is also the world's largest roll-on/roll-off (RoRo) ship, with a capacity of 9,200 vehicles.
The Lurline is the largest roll-on/roll-off (Ro-Ro) ship ever built in the United States. It was built by NASSCO General Dynamics in San Diego. It has room for approximately 500 vehicles.
BYD is also the world's largest producer of electric vehicles (EVs). In 2023, BYD manufactured 3 million EVs, while Tesla produced 1.8 million.
The Danyang-Kunshan Grand Bridge in China holds the record for the world's longest bridge (102 miles).
The Lake Pontchartrain Causeway in Louisiana is the longest bridge in the United States (24 miles).
China’s Tiangong Space Station was launched on April 29, 2021. The International Space Station is now 26 years old (launched in 1998) and the USA is currently working on plans for its deorbit and disposal. As of July 2024, NASA has reportedly not seen any viable proposals for reuse of the ISS or parts of it.
Unmanned Chinese spacecraft Chang'e 4 successfully landed on the moon's far side in 2019, and another craft landed in 2020 (Chang'e 5), returning with thousands of grams of samples.
The last manned USA moon mission was Apollo 17 in 1972. No US mission, manned or unmanned, has returned to the moon until just last year, with the Odysseus lander touching down in February of 2022. The next NASA moon landing is tentatively planned for 2027, but a negative safety report that has recently come out from the Aerospace Safety Advisory Panel (ASAP) may put that timing into question.
The Chengdu J-36 and the Shenyang J-XX/J-50 are two sixth-generation combat aircraft prototypes that have already been shown to be complete and ready for action in December of 2024
The United States Air Force (USAF) and United States Navy (USN) are expected to field their first sixth-generation fighters in the 2030s.
I could go on.
Do you still feel China has any difficulty producing top-notch technology?
So, my question now is: is the economy of the western world losing its hegemony on technology?
Of course, I don’t presume have the tools or knowledge to properly and fully answer that question in a meaningful way. I can only express my own opinions on the matter.
And my opinion is we might be.
There are many factors I feel are closely tied to why the western world is so blatantly being left behind in a field we quite literally once wrote the book on.
One of the first, and most important, is education.
There is a distinct, palpable, difference between the systems of higher education (i.e. universities and colleges) we have in western countries compared with those in China. I will focus on those within the USA, as those are the ones I have the most personal experience with.
I’ve never been to China myself, but after living for 20+ years in California’s Silicon Valley, where 14% of the workforce in highly technical occupations is made up of people of Chinese origins, has given me some very consistent insight over the years.
Let’s start with some numbers.
In 2023, China had 10.5M undergraduates, 5 times more than the USA (this second number also includes any international students currently studying in the USA).
In 2030, the projection is that 27% of the world's undergraduates will come from China. Only 8% will be from the USA.
The United States (340 million inhabitants) graduates roughly 70,000 undergraduate engineers annually, whereas China graduates 600,000, and India 350,000.
In 2023, the university enrollment rate in China was 60.2% according to the Ministry of Education, meaning 47 million mainland Chinese students (out of a population of 1.4 billion) were at that time enrolled in 4-year university and college degree-granting programs.
Entry into Chinese universities is intended to be meritocratic, depending only on the result of the Gaokao entrance examination. Entry is reportedly not influenced or determined by sporting activities, extracurricular programs, donations, or alumni parents and siblings.
In 2017, China surpassed the USA with the highest number of scientific publications. To this day, the vast majority of research papers on AI come from Chinese academics.
For undergraduate studies at a public university in China, you are expected to pay anywhere between $2,000 to $10,000 a year, depending on the university and major. For private universities, the cost fluctuates up to $30,000 a year.
The average cost in the USA for a public college is $40,000 a year. It goes up to $70,000 a year for private colleges.
A significant conclusion comes to mind based on the data above: America’s higher-ed system has deeply rooted issues tied to costs and quality.
I have also seen some major attitude shifts regarding the idea of college as a whole in recent years, and not in a positive direction. But I’ll get into that in a moment.
The cost of going to college is a very tangible issue that we could, if we wanted to, act upon. My obvious top suggestion would be to go back to the pre-Reagan era when public colleges in America were free. (Yes, college in the USA used to be free, can you believe that?).
If that sounds too extreme, the least we could do is to dramatically reduce the tuition costs to public higher-ed institution to increase their rate of enrollment. Once folks don’t have to get into hundreds of thousands of dollars’ worth of debt before they even turn 20, I’m sure college will become a much more appealing option to some.
A great first step to ease into this process, would be to increase and extend all financial aid in existence today. There are several universities across the country who are starting to implement programs such as this. One such example is New York University, a private college, mind you, where any students with a family gross income below $100,000 get free tuition.
The statistics say the US Economy needs about 400,000 engineers a year to meet the current needs of the market, and the US college system (public and private) can provide only 70,000 of them.
Matching this number should be the top priority for any government if we want to regain the ground we have lost. The era of corporations massively hiring internationally is over, mostly due to the current administration’s immigration policies. On the other hand, hiring talents locally in other countries and letting them work remotely is very hard and many companies have failed in such experiments. That kind of work structure requires a very carefully planned scaling of project management, chief amongst which is giving remote offices a proper level of independence and consistently assigning them projects of an adequate importance. They can’t simply be treated as specialized manpower and be assigned work that a company’s US-based employees might just not want to do. The low costs of the workforce is also no longer a valid excuse since the salaries of highly qualified workers have exponentially increased, even in notoriously low-wage labor markets like India.
Faced with these conditions, these international talents will eventually leave if they feel their contributions are not properly appreciated or if the complexity of the challenges assigned to them is not at par with their expectations. This might sound pedantic to say, but these are people, not parts of an assembly line. They have their own career expectations and goals, and if they are not satisfied, they will simply leave. The quality of their education and their expertise guarantees that they will not lack in better opportunities elsewhere.
Speaking of quality of schooling, that is another item in which the west is unfortunately clearly falling behind on in comparison to Chinese education.
To offset that, we should start teaching specialized and advanced classes in the first few years of college and get rid of General Education (GE) classes, which are essentially a repetition of high school material.
GE classes were originally introduced to equalize the knowledge base of students coming to college from different high schools from all over the country. It is well known that the preparation level is not homogeneous among high schools: even if there are general federal guidelines regarding education, each individual US state has discretion over the goals and budget for its own education system.
This is not a US-only specific problem. Even in Italy, where all public education is directly managed at a national level, there are still differences in student preparation depending on which high school they attend. The difference is that in Italy, universities just assume everyone comes with a specific baseline of knowledge and it is the responsibility of each individual student to catch up if they have any gaps. “Leveling” (GE) classes do not exist.
The style and content of exams should also be rethought to make them less academic and more shaped as mini projects. With the current format of the American educational system, we see students aiming just to pass the next test and no interest in actually absorbing the learning material as they should. A lot of the time they do not remember a thing of what they studied just a few weeks after the test, since their goal is now to pass the next one. Don’t even get me started about how many of them have now started to use ChatGPT to do their homework or even the test! Having an exam that forces you to apply the concepts you just learned could be a good direction to guide the students to tackle exams more seriously, and also would be more useful for them in the long run.
Now, let’s talk about the attitude problem.
In recent years, I’ve noticed an increasingly prevalent movement among teenagers in several western countries (namely the USA and Italy) where being a good student and having good grades is now seen as a negative mark on their social profile. To facilitate their acceptance in the pack, they prefer to look like someone who doesn’t care about getting a good education and great grades, not to mention the general belief in the usefulness of a college degree as a whole is in decline. Again, this is not solely an American phenomenon, but it is prevalent, and it is getting louder. The diminished role of parents (in a positive sense) in their children’s education is also a huge issue in my mind, but that is a topic for another time.
Dictating what the newest popular “fad” is or what makes a teenager “popular” in the eyes of their peers on any particular day is an impossible task. However, we, as a society, definitely have some things we do have the power to address and correct.
In US universities, there is this nice, at least in theory, mechanism of continuous student feedback regarding their professors and class material. This data is typically collected with surveys or questionnaires at the end of every term. American colleges take this feedback very seriously, often informing hiring and firing practices, as well as instructor salary, while in Italy they are just … informative.
I understand the noble reasoning behind this, but I’ve also faced situations where teachers are pushed into a corner with ridiculous complaints from students. Some have even lost their jobs thanks to the results of these kinds of surveys. The level of arrogance some students have, often feeling disproportionately empowered by these tools, is unbelievable, and they certainly don’t hesitate to flex this power. Unfortunately, they don’t always do so in good faith and often the results, as studies have shown, are inherently biased.
This perverse game is particularly true in private colleges where the admin staff is usually more indulgent in listening to and protecting the “paying customer”, the student. I truly believe giving the students the chance to send feedback to the institution is a positive thing, but we should also give the proper weight to these notes and only extract from them an appropriate level of suggestions for improvement and not allow them to be used as fuel for personal disputes between students and professors.
The issue here is also that US colleges kick students out if they do not get and maintain a proper minimum GPA. This policy is great at not having eternal repeaters and enforcing student commitment to their education. In Italy, a bachelor’s degree should take about three years, but most students graduate within an average of 5.1 years. This even happens with master’s programs, where the average graduation time is almost three years, when most programs are only two years long. On the other hand, colleges need graduates as they pay tuition, and they need students to have good grades and the reputation, admissions rate, and graduation rate of a college strongly affects its enrollment numbers. This means admin staff are always having to contend with these two contradictory goals, and sometimes the budget inevitably wins.
Let me add some personal experience based on the years my daughter spent as a student in the American education system.
On average, a Chinese student works very, very hard.
It has become a stereotype at this point, but nobody can deny that there is a deep historical truth to it.
The pressure these students get from their families is often insane. There were multiple times I was witnessed moments during my daughter’s schooling where the parents of Chinese students complained with their teachers about the class program (mostly in relation to math), asking for harder content as the current curriculum didn’t intellectually stimulate their kids enough.
Several times, my daughter noticed Chinese kids in her classes that weren’t actually paying attention to what was being said in class and instead were reading advanced content or going ahead in the textbook as they already knew the topic. On top of that, many would go to extra private tutoring sessions on the weekends.
Of course, if we consider their quality of life as teenagers, we may express some concern, but this is their reality.
To summarize, we have a mix of class content issues, mostly in all the fields of STEM, as well as issues on the school organization and management level.
One way to start bridging the gap would be to enforce merit and competence as admission criteria. Let’s stop with the sport scholarships in the way they work today. I understand that very often sports scholarships are the only chance for students coming from disadvantaged realities to get access to a college education in America, but we can help those kinds of students with stronger financial aid programs without discounting the kids’ merit. We should enforce students merit guidelines independent of their athletic achievements.
The summit of technological achievement cannot be attained just by having large capitals available to be invested. The human factor is the essential ingredient to progress and innovation. Our future researchers, scientists, and engineers need the proper education and resources to succeed.
Our population numbers are already against us. Census reports are clear: western communities are shrinking, Asian and African communities are booming. We need to take care of the best and brightest people we have and help them achieve their dreams. In the long run, this will not only have individual benefits, but it will also improve our countries as a whole.
So, to go back to the question: Are we reliving another Sputnik moment?
No. No we're not.
Russia in 1957 surprised the world. With Sputnik, they revealed an achievement people back then with no access to the modern ways of instant communication could have possibly known about until it actually happened.
China has been shouting its massive achievements and quality of its researchers and scientists from the rooftops for decades.
It was us, the Western world, that failed to listen, and if we don't start doing so now, learn from our mistakes, and stop underestimating this global superpower, there might truly be no way to catch up again, and who knows what that would mean.