Artificial intelligence is transforming the entire pipeline from college to the workforce: from tests and grades to job applications and entry-level work.
Honestly I think it’s even worse than this because so many of them came of age at a time when it’s been hard to cultivate attention and critical thinking. This past semester I worked as an adjunct writing instructor for a graduate school class at my alma mater. It was a demanding course, with a lot of reading (including philosophical texts and legal decisions). I think most students took it only because it’s required.
While a few of the students seemed engaged, quite a few seemed unable to grasp much of the material and to lack the wherewithal to structure or even understand a complex argument. It seemed as if their prior education hadn’t prepared them well. It was a frustrating experience for me - those “students don’t or can’t read books” posts really resonated - but I feel terrible for them and worry about their prospects.
I had this exact experience teaching a course at a top liberal arts college. The course was structured as a discussion-based seminar because it had 16 students and it was at... a top liberal arts college. One would assume that is the experience students had opted into. And yet only 3-4 students regularly contributed (despite participation being 10% of their grade). Students just scrolled their laptops during class. Moving forward, I am banning screens from the classroom.
But I think you're getting at part of the problem. The students aren't able to read the dense readings -- or at least, to attempt to read them and struggle -- and so they check out. There's only so much college professors can do when students fail upwards. I routinely hear about the pressure for K-12 teachers to just pass students and grade inflation has hit high schools hard.
For another perspective, I help with undergraduate recruiting in a white-collar role that a lot of students from top colleges fall into (banking, consulting, finance). We used to primarily hire from top schools (“targets”) but also leave the pipeline open to cold applications from other schools. The quality of cold applications was usually lower, but there were always a few standouts.
But the sheer volume of non-target applications, influenced by AI, has led us to drastically cut back non-target hiring. We literally can’t read them all. And so now we hire even more from our target schools, which means that the decision that most affects an applicants post-grad placement is whether an admissions counselor let them into Columbia or Duke at age 17. Our current state is an equilibrium, but it’s not a good or better one than what we had before.
First, I am a huge fan of your work—Plain English, The Atlantic, Hit Makers, and now I am very excited to see you on Substack.
I am one of these unemployed new grads, and have been so for around a year with no progress. I have a graduate engineering degree and studied writing as well. I worked hard to do well in school, and am just old enough to have missed the AI wave as a student. I have applied to well over a hundred jobs at this point, and am finding the process not only fruitless, but entirely bleak. Your reporting is spot on, and I thank you for that! I think there are a few extra angles that have yet to be discussed much, many of which align with your past coverage of adult loneliness and the attention economy:
1. Covid eliminated the backdoor—I was a junior in college when the pandemic struck. At this point, I and many of my classmates gravitated toward existing social and academic circles rather than reaping the benefit of network expansion that college typically offers. There was little cross pollination within remote elective courses, as students were less willing to take risks and branch outside of their major. This, combined with the (understandable) preference of many students to take Zoom classes with their cameras off, invest less time in communal office hours and group projects due to the option of pass/fail grading, and the general depletion of the college town ecosystem, cauterized social and professional networks. When finding a job through conventional routes such as hiring sites seems impossible, the option to reach out to friends/connections of friends for help finding a role through the backdoor is less available. The ability to ask for a favor seems inaccessible.
2. Loneliness—The isolation of unemployment is stifling to the point of inertia. I do not interact with many people, certainly not many peers, since moving back home. This is mainly a symptom of being in an area with few young people, but also of the effects that job searching poses on confidence, sanity, and overall appraisal of life. It is very difficult to digest the reality that I provide little economic value, at least at this moment. There is a lot of evidence that the inability to provide for oneself/one’s family is a driver of suicide in men especially. Couple this issue with loneliness, online entrenchment, and constant rejection, and you land on a recipe for a massive mental health antagonist. This not only stifles wellbeing at large, but also presents an impediment to the job search process: When you become hopeless with little relief, there is diminished incentive to continue your pursuit of a better life, as the negative opportunity cost is largely driven (at that point) by the effort and emotional input of applying for jobs vs. a consistently low probability of getting hired regardless of how much you try. In other words, the effort of applying for a job seems moot, and therefore unattractive. Which ultimately opens up a pit of inertia-driven despair, which breeds low self-esteem/a lack of incentive to invest in life. This unemployment crisis will, among many other externalities, create a huge bottleneck in social skills—not only are young adults not exposed to heterogeneous groups within a workplace, but we are also too discouraged by life to go out into the world an socialize with peers.
3. Attention economy—This one may be a little difficult to articulate, but I think it’s a very important point. The internet has acclimated us to strive for perfection and immediate relief/diversion. Algorithms are seamlessly tailored to our tastes, and anything less than ideal can be blocked. Our tolerance of ideological differences and ability to exercise critical thinking has lowered, while our appetite for quick bits of stimulation has become insatiable. Neither of which are favorable for a workplace. My generation is accustomed to our needs being completely satisfied online, and I believe this mindset trickles into our expectation of work as well. This manifests in two related schools of thought: Either no job except our dream job will be good enough for us, or we malign capitalism and the premise of work as an unequivocally evil force to be shunned at large. I know this doesn’t completely square with the fact that we still are applying to jobs, many of which I’m sure are perceived to be far from ideal (and I’m guessing socioeconomic background plays into this significantly as well). But I do think that these beliefs, combined with the destitution of today’s job market, are driving us to reconsider the value that work brings to our lives.
I think one big question to ask as a result is what kind of job openings are young people applying to? Do they feel (socially, ideologically) comfortable settling for a position that is below their level of education/expectations, and therefore not in conformance with their appetite for perfection? This is by no means an indictment of those who have rightfully high standards (I am guilty of this myself). But a generation that is accustomed to being constantly indulged by technology may have a predilection to expect the same from their job. In a moment where it is unlikely to get the job we think we deserve, are we motivated to optimize our chances of entering the workforce by pursuing a temporarily less glamorous, less than ideal path in the meantime? Or do we bury our heads in the sand and not “read the economic room” but rather continue to dedicate much of our job search to positions outside of the scope of reality/probability? I honestly don’t know the answer and would love to see data.
For example, both of my degrees are from one of the universities whose career office you spoke to for this piece. I truly respect the institution, and believe they do great work in accommodating their alumni network. I loved my time there. But their job board for alumni clearly is an unfortunate microcosm of the current state of the job market—among a small population of jobs for marketing managers and project engineers is a deluge of postings for cooks, baby/pet sitters, and construction workers. It’s frustrating to see my university, which I paid close to $200K of tuition to, to have titles such as Sandwich Artist earnestly posted as an option for its graduates to pursue. And while this is reality and I understand that I’m not entitled to anything, it’s tough to swallow my pride and embrace it.
If I could ask for anything, it would be for the social impact of work to be taken seriously as an incentive to reinvigorate the job market, or at least for the positives of AI to be weighed earnestly against the psychological impact of supplanting human labor. AI and social media are a scourge on developing minds. Unemployment and the isolation it facilitates only exacerbates their effects. A rich utopia is no fun if everyone is adrift and depressed.
I too am another unemployed grad like you. I think your first point is essential and that your second two points feed into the first.
I had a conversation with a WSJ reporter two weeks ago on this exact topic—where they asked for a bit of my thoughts on the new grad gap.
First I said there are two factors driving this storyline.
A) Writing about AI gets clicks and media today is highly dependent on clicks.
B) Corporations would love to tell you they are not hiring because of increases in productivity and efficiency (AI). They would not like to tell you they are not hiring because they are deeply unsure about the direction of the economy (Unclear Growth Prospects).
This is to say I think that there is a macro trend of cautious hiring as a result of the extensive uncertainty in the economy over the past few years. And the chart here shows this. The trend starts way before AI is talked about. This is then exacerbated for young college grads by their lack of social capital.
The marketing around college says it’s only about human capital (learning stuff) but college is also about social capital (meeting people). COVID decreased the first and decimated the latter.
We, and young people in particular, are dealing with those consequences—a direct result of sacrificing the youth in the hopes of saving the elderly. A political tradeoff that has not been without consequence for democrats. But I digress.
As you say without those weak ties, it is very hard to ask for a favor. And since applying through portals is a waste of time, you need a reference. Which is to say you need a favor but there are only so many favors to ask.
And without cheap housing in economically vibrant places—moving and figuring it out becomes a much harder economic calculus.
So, no, I would not chalk it up to AI, but look a little deeper, and of course, in the spirit of Mr. Thompson—talk about housing! But to my disappointment I have yet to see the WSJ reporter or anyone else dig deeper and talk about the social aspect in any serious way because I do think it is a bit more relevant.
I'm a political normie in most respects, but AI is the one issue I would never deliberately raise in polite conversation. I think it's going to make the next 5-20 years profoundly weird -- and that's assuming it DOESN'T turn is into gray goo. We're on the precipice of sea change after sea change, and real-world discourse is nowhere close to catching on.
With respect to work, I think a couple of trends are likely to emerge as jobs become harder and harder for new grads to find:
1) Entreprenuerialism is going to be rewarded handsomely. In a world where the conveyor belt to financial freedom no longer exists (or at least is murkier than it once was), lots of bright young people are going to have to build their own.
2) Somewhat downstream of #1 -- we're going to see more young people in blue-collar industries that LLMs have yet to touch. Think more laundromat owners, construction project management, and the like.
I agree blue collar jobs are going to be on the rise the problem with that I see is most of those jobs don’t pay people enough at this point to afford living in the U.S at this point. So unless the cost of living decreases (which we know it never will) how will kids who choose that path pay for homes, kids, cars etc?
Oh for sure -- not insinuating it doesn't! Just name-checking it as an industry that becomes more attractive in the next however many years relative to, say, law.
As a lawyer (not currently practicing) who graduated during the previous millennium, I will observe that the legal profession has long had many unattractive qualities 😉
My colleague at the Upjohn Institute for Employment Research, labor economist Brad Hershbein, has a Bluesky thread which argues that if you look at employment rates and hiring rates, the problem with recent college grads is NOT some AI trend interfering with their hiring. https://bsky.app/profile/bradhershbein.bsky.social/post/3lsjdo7pipk25
Rather, if you look at employment rates, what you see is a GENERAL weakening of the labor market, which affects young people in particular, not just those with a BA, particularly when compared with the 2021-2022 period, which was unusual.
In other words, the labor market problem is the economic uncertainty that is depressing new hires, not AI.
botched my first comment. anyways, what are your thoughts on this shifting the impetus of hiring, learning, etc. back onto "those in charge". for example, back in the day, you could be sure that a low baseline of some knowledge made you (relatively) smart and could get a job with a firm handshake (hyperbole, but you know what i mean). then, the burden of knowledge made it such that students had to study more to be productive members of society (higher age for patents, entry level roles having more difficult prerequisites, etc.) and application processes became so arduous by having nonstandard application processes and lengthy interview processes. now, shouldn't universities have to use their bloated endowments to hire/support dedicated lecturers and employ some better epistemology than reusing old midterms and assigning lengthy homeworks? and companies will have to employ larger hr departments to dedicate resources to finding "who actually knows what" as opposed to CV filters and lengthy take-home assignments?
I’m a millennial father of two young boys. When I read articles like this one and others about how much AI has infiltrated the education system and hiring practices I’m extremely worried I’m not going to be able to help them or even give them sold advice.
I don’t use AI for my job or in everyday life, I just haven’t found a reason to. Reading, writing, and communicating were all skills I developed in college that have benefited me in all my professional careers.
When it’s time for high school and college I really hope the my boys and students still develop these skills even in a generation dominated by AI because I just don’t understand how you go through life without thee core skills.
Mark Granovetter wrote a research paper titled "Strength of Weak Ties" in 1973. His research showed that you are 70% more likely to find a job (or any sort of diverse thinking or novel ideas) through your weak ties. Your strong ties are your friends and family. Your weak ties are your neighbors, a friend from college you haven't seen in years, or an acquaintance.
They replicated the results and gleaned a bit more data on the type of ties that are best. All of this is to say, regardless of the existence of artificial intelligence used in the screening process, you are still most likely to find a job through people you know and telling them what types of opportunities you're seeking.
Trust me, I'm an introvert with social anxiety. I've looked for every other possible way. If I could hide behind AI, I would. But in the future, humans will still do business with humans, and people will hire people.
"“I’ve had students accused of using AI in the interview process,” one college career executive told me. “The student swears to me that they weren’t cheating, but in a virtual interview when they have access to a computer, it’s hard for the recruiter to know.”"
If they're asking you to produce written content, and you produce something that is of a certain quality, why does it matter if they used AI? Isn't that just a tool you use to improve your skills? I'm not sure how this is "cheating" any more than accountants "cheat" by using calculators.
Derek. — I’ve been teaching internet media history at nyu for 16 years and I greatly appreciated the Atlantic piece. On the follow up — a couple of thoughts. Career centers on campuses are often not the source of first jobs for college students. Second, given all that is happening, where are college students supposed to turn for how to navigate the very real impact that students have gelt this spring and summer.
Much more can be said about about how this is specifically impacting entry level software engineers. AI already is taking over the type of work these individuals would take on, and it's not clear if any company has a strategy to still hire inexperienced, cheap engineers, and still develop them to be real contributors.
We just received a note from our CEO at the relatively mature tech startup I work at about operating lean as we attempt to reach profitability. AI was explicitly mentioned as a substitute for hiring more people to our teams as we try to accomplish more with fewer people.
What I wonder is: will this increase in efficiency of those who wield AI tools effectively yield higher wages? Seems like it should in theory. I'm not holding my breath in the short run, kinda glad to have a job.
This isn't a new trend; it's just an acceleration in an existing arms race.
Automated HR screening has been a thing for decades now, as have mass application tools. The difference is just that instead of manually tailoring a resume for a job description (I am old enough to remember the trick of copy/pasting the JD and then minimizing & whiting out the font) the AI will do the automating for you.
Abstractly, the problem of resume review is of extracting a signal of candidate quality. That signal has become increasingly low quality. The various automated systems are a response to try and filter, but of course all filtering can be gamed and so the cycle continues.
Honestly at least in my industry everyone knows just blind applying for jobs is a pointless exercise; your probability of getting through the thousands of resumes that are going through the automated filter is very close to zero. Everything is about getting a high-quality referral. Which of course makes life hard for people who don't have a professional network!
Great read, Derek. Anecdotally - as an ancient Gen Z (27) it blows my mind to see the shift in experience from when I applied/got my job postgrad (5 years ago, how already?) to my friends searching for work now (1st/2nd/3rd jobs). AI, similar to COVID, feels like an inflection point of pre and post.
It brings to mind a thinker that I respect, Andy Crouch, who talks about technology as human's quest for magic (instant, effortless power), and how "Modern technology represents, in some ways, the most successful attempts we have yet seen of getting the world to do magic on our behalf."
But it comes with a cost on what it means to be human. Which raises the question for me- are we actually getting what we bargained for?
From my observation, whether it's online dating or applying to hundreds of job postings, we're increasing our pools/options but it feels like we're in a time of heightened scarcity…not abundance from the standpoint of human experience (relationships, face to face interactions both personally and professional, a sense of community, finding employment).
Honestly I think it’s even worse than this because so many of them came of age at a time when it’s been hard to cultivate attention and critical thinking. This past semester I worked as an adjunct writing instructor for a graduate school class at my alma mater. It was a demanding course, with a lot of reading (including philosophical texts and legal decisions). I think most students took it only because it’s required.
While a few of the students seemed engaged, quite a few seemed unable to grasp much of the material and to lack the wherewithal to structure or even understand a complex argument. It seemed as if their prior education hadn’t prepared them well. It was a frustrating experience for me - those “students don’t or can’t read books” posts really resonated - but I feel terrible for them and worry about their prospects.
I had this exact experience teaching a course at a top liberal arts college. The course was structured as a discussion-based seminar because it had 16 students and it was at... a top liberal arts college. One would assume that is the experience students had opted into. And yet only 3-4 students regularly contributed (despite participation being 10% of their grade). Students just scrolled their laptops during class. Moving forward, I am banning screens from the classroom.
But I think you're getting at part of the problem. The students aren't able to read the dense readings -- or at least, to attempt to read them and struggle -- and so they check out. There's only so much college professors can do when students fail upwards. I routinely hear about the pressure for K-12 teachers to just pass students and grade inflation has hit high schools hard.
For another perspective, I help with undergraduate recruiting in a white-collar role that a lot of students from top colleges fall into (banking, consulting, finance). We used to primarily hire from top schools (“targets”) but also leave the pipeline open to cold applications from other schools. The quality of cold applications was usually lower, but there were always a few standouts.
But the sheer volume of non-target applications, influenced by AI, has led us to drastically cut back non-target hiring. We literally can’t read them all. And so now we hire even more from our target schools, which means that the decision that most affects an applicants post-grad placement is whether an admissions counselor let them into Columbia or Duke at age 17. Our current state is an equilibrium, but it’s not a good or better one than what we had before.
Derek,
First, I am a huge fan of your work—Plain English, The Atlantic, Hit Makers, and now I am very excited to see you on Substack.
I am one of these unemployed new grads, and have been so for around a year with no progress. I have a graduate engineering degree and studied writing as well. I worked hard to do well in school, and am just old enough to have missed the AI wave as a student. I have applied to well over a hundred jobs at this point, and am finding the process not only fruitless, but entirely bleak. Your reporting is spot on, and I thank you for that! I think there are a few extra angles that have yet to be discussed much, many of which align with your past coverage of adult loneliness and the attention economy:
1. Covid eliminated the backdoor—I was a junior in college when the pandemic struck. At this point, I and many of my classmates gravitated toward existing social and academic circles rather than reaping the benefit of network expansion that college typically offers. There was little cross pollination within remote elective courses, as students were less willing to take risks and branch outside of their major. This, combined with the (understandable) preference of many students to take Zoom classes with their cameras off, invest less time in communal office hours and group projects due to the option of pass/fail grading, and the general depletion of the college town ecosystem, cauterized social and professional networks. When finding a job through conventional routes such as hiring sites seems impossible, the option to reach out to friends/connections of friends for help finding a role through the backdoor is less available. The ability to ask for a favor seems inaccessible.
2. Loneliness—The isolation of unemployment is stifling to the point of inertia. I do not interact with many people, certainly not many peers, since moving back home. This is mainly a symptom of being in an area with few young people, but also of the effects that job searching poses on confidence, sanity, and overall appraisal of life. It is very difficult to digest the reality that I provide little economic value, at least at this moment. There is a lot of evidence that the inability to provide for oneself/one’s family is a driver of suicide in men especially. Couple this issue with loneliness, online entrenchment, and constant rejection, and you land on a recipe for a massive mental health antagonist. This not only stifles wellbeing at large, but also presents an impediment to the job search process: When you become hopeless with little relief, there is diminished incentive to continue your pursuit of a better life, as the negative opportunity cost is largely driven (at that point) by the effort and emotional input of applying for jobs vs. a consistently low probability of getting hired regardless of how much you try. In other words, the effort of applying for a job seems moot, and therefore unattractive. Which ultimately opens up a pit of inertia-driven despair, which breeds low self-esteem/a lack of incentive to invest in life. This unemployment crisis will, among many other externalities, create a huge bottleneck in social skills—not only are young adults not exposed to heterogeneous groups within a workplace, but we are also too discouraged by life to go out into the world an socialize with peers.
3. Attention economy—This one may be a little difficult to articulate, but I think it’s a very important point. The internet has acclimated us to strive for perfection and immediate relief/diversion. Algorithms are seamlessly tailored to our tastes, and anything less than ideal can be blocked. Our tolerance of ideological differences and ability to exercise critical thinking has lowered, while our appetite for quick bits of stimulation has become insatiable. Neither of which are favorable for a workplace. My generation is accustomed to our needs being completely satisfied online, and I believe this mindset trickles into our expectation of work as well. This manifests in two related schools of thought: Either no job except our dream job will be good enough for us, or we malign capitalism and the premise of work as an unequivocally evil force to be shunned at large. I know this doesn’t completely square with the fact that we still are applying to jobs, many of which I’m sure are perceived to be far from ideal (and I’m guessing socioeconomic background plays into this significantly as well). But I do think that these beliefs, combined with the destitution of today’s job market, are driving us to reconsider the value that work brings to our lives.
I think one big question to ask as a result is what kind of job openings are young people applying to? Do they feel (socially, ideologically) comfortable settling for a position that is below their level of education/expectations, and therefore not in conformance with their appetite for perfection? This is by no means an indictment of those who have rightfully high standards (I am guilty of this myself). But a generation that is accustomed to being constantly indulged by technology may have a predilection to expect the same from their job. In a moment where it is unlikely to get the job we think we deserve, are we motivated to optimize our chances of entering the workforce by pursuing a temporarily less glamorous, less than ideal path in the meantime? Or do we bury our heads in the sand and not “read the economic room” but rather continue to dedicate much of our job search to positions outside of the scope of reality/probability? I honestly don’t know the answer and would love to see data.
For example, both of my degrees are from one of the universities whose career office you spoke to for this piece. I truly respect the institution, and believe they do great work in accommodating their alumni network. I loved my time there. But their job board for alumni clearly is an unfortunate microcosm of the current state of the job market—among a small population of jobs for marketing managers and project engineers is a deluge of postings for cooks, baby/pet sitters, and construction workers. It’s frustrating to see my university, which I paid close to $200K of tuition to, to have titles such as Sandwich Artist earnestly posted as an option for its graduates to pursue. And while this is reality and I understand that I’m not entitled to anything, it’s tough to swallow my pride and embrace it.
If I could ask for anything, it would be for the social impact of work to be taken seriously as an incentive to reinvigorate the job market, or at least for the positives of AI to be weighed earnestly against the psychological impact of supplanting human labor. AI and social media are a scourge on developing minds. Unemployment and the isolation it facilitates only exacerbates their effects. A rich utopia is no fun if everyone is adrift and depressed.
I too am another unemployed grad like you. I think your first point is essential and that your second two points feed into the first.
I had a conversation with a WSJ reporter two weeks ago on this exact topic—where they asked for a bit of my thoughts on the new grad gap.
First I said there are two factors driving this storyline.
A) Writing about AI gets clicks and media today is highly dependent on clicks.
B) Corporations would love to tell you they are not hiring because of increases in productivity and efficiency (AI). They would not like to tell you they are not hiring because they are deeply unsure about the direction of the economy (Unclear Growth Prospects).
This is to say I think that there is a macro trend of cautious hiring as a result of the extensive uncertainty in the economy over the past few years. And the chart here shows this. The trend starts way before AI is talked about. This is then exacerbated for young college grads by their lack of social capital.
The marketing around college says it’s only about human capital (learning stuff) but college is also about social capital (meeting people). COVID decreased the first and decimated the latter.
We, and young people in particular, are dealing with those consequences—a direct result of sacrificing the youth in the hopes of saving the elderly. A political tradeoff that has not been without consequence for democrats. But I digress.
As you say without those weak ties, it is very hard to ask for a favor. And since applying through portals is a waste of time, you need a reference. Which is to say you need a favor but there are only so many favors to ask.
And without cheap housing in economically vibrant places—moving and figuring it out becomes a much harder economic calculus.
So, no, I would not chalk it up to AI, but look a little deeper, and of course, in the spirit of Mr. Thompson—talk about housing! But to my disappointment I have yet to see the WSJ reporter or anyone else dig deeper and talk about the social aspect in any serious way because I do think it is a bit more relevant.
I'm a political normie in most respects, but AI is the one issue I would never deliberately raise in polite conversation. I think it's going to make the next 5-20 years profoundly weird -- and that's assuming it DOESN'T turn is into gray goo. We're on the precipice of sea change after sea change, and real-world discourse is nowhere close to catching on.
With respect to work, I think a couple of trends are likely to emerge as jobs become harder and harder for new grads to find:
1) Entreprenuerialism is going to be rewarded handsomely. In a world where the conveyor belt to financial freedom no longer exists (or at least is murkier than it once was), lots of bright young people are going to have to build their own.
2) Somewhat downstream of #1 -- we're going to see more young people in blue-collar industries that LLMs have yet to touch. Think more laundromat owners, construction project management, and the like.
I agree blue collar jobs are going to be on the rise the problem with that I see is most of those jobs don’t pay people enough at this point to afford living in the U.S at this point. So unless the cost of living decreases (which we know it never will) how will kids who choose that path pay for homes, kids, cars etc?
Construction management is hard and requires *a lot* of skills
Oh for sure -- not insinuating it doesn't! Just name-checking it as an industry that becomes more attractive in the next however many years relative to, say, law.
As a lawyer (not currently practicing) who graduated during the previous millennium, I will observe that the legal profession has long had many unattractive qualities 😉
My colleague at the Upjohn Institute for Employment Research, labor economist Brad Hershbein, has a Bluesky thread which argues that if you look at employment rates and hiring rates, the problem with recent college grads is NOT some AI trend interfering with their hiring. https://bsky.app/profile/bradhershbein.bsky.social/post/3lsjdo7pipk25
Rather, if you look at employment rates, what you see is a GENERAL weakening of the labor market, which affects young people in particular, not just those with a BA, particularly when compared with the 2021-2022 period, which was unusual.
In other words, the labor market problem is the economic uncertainty that is depressing new hires, not AI.
botched my first comment. anyways, what are your thoughts on this shifting the impetus of hiring, learning, etc. back onto "those in charge". for example, back in the day, you could be sure that a low baseline of some knowledge made you (relatively) smart and could get a job with a firm handshake (hyperbole, but you know what i mean). then, the burden of knowledge made it such that students had to study more to be productive members of society (higher age for patents, entry level roles having more difficult prerequisites, etc.) and application processes became so arduous by having nonstandard application processes and lengthy interview processes. now, shouldn't universities have to use their bloated endowments to hire/support dedicated lecturers and employ some better epistemology than reusing old midterms and assigning lengthy homeworks? and companies will have to employ larger hr departments to dedicate resources to finding "who actually knows what" as opposed to CV filters and lengthy take-home assignments?
I’m a millennial father of two young boys. When I read articles like this one and others about how much AI has infiltrated the education system and hiring practices I’m extremely worried I’m not going to be able to help them or even give them sold advice.
I don’t use AI for my job or in everyday life, I just haven’t found a reason to. Reading, writing, and communicating were all skills I developed in college that have benefited me in all my professional careers.
When it’s time for high school and college I really hope the my boys and students still develop these skills even in a generation dominated by AI because I just don’t understand how you go through life without thee core skills.
Mark Granovetter wrote a research paper titled "Strength of Weak Ties" in 1973. His research showed that you are 70% more likely to find a job (or any sort of diverse thinking or novel ideas) through your weak ties. Your strong ties are your friends and family. Your weak ties are your neighbors, a friend from college you haven't seen in years, or an acquaintance.
Stanford University ran the same study again in 2022 using LinkedIn. https://news.stanford.edu/stories/2022/09/real-strength-weak-ties
They replicated the results and gleaned a bit more data on the type of ties that are best. All of this is to say, regardless of the existence of artificial intelligence used in the screening process, you are still most likely to find a job through people you know and telling them what types of opportunities you're seeking.
Trust me, I'm an introvert with social anxiety. I've looked for every other possible way. If I could hide behind AI, I would. But in the future, humans will still do business with humans, and people will hire people.
"“I’ve had students accused of using AI in the interview process,” one college career executive told me. “The student swears to me that they weren’t cheating, but in a virtual interview when they have access to a computer, it’s hard for the recruiter to know.”"
If they're asking you to produce written content, and you produce something that is of a certain quality, why does it matter if they used AI? Isn't that just a tool you use to improve your skills? I'm not sure how this is "cheating" any more than accountants "cheat" by using calculators.
Derek. — I’ve been teaching internet media history at nyu for 16 years and I greatly appreciated the Atlantic piece. On the follow up — a couple of thoughts. Career centers on campuses are often not the source of first jobs for college students. Second, given all that is happening, where are college students supposed to turn for how to navigate the very real impact that students have gelt this spring and summer.
Much more can be said about about how this is specifically impacting entry level software engineers. AI already is taking over the type of work these individuals would take on, and it's not clear if any company has a strategy to still hire inexperienced, cheap engineers, and still develop them to be real contributors.
We just received a note from our CEO at the relatively mature tech startup I work at about operating lean as we attempt to reach profitability. AI was explicitly mentioned as a substitute for hiring more people to our teams as we try to accomplish more with fewer people.
What I wonder is: will this increase in efficiency of those who wield AI tools effectively yield higher wages? Seems like it should in theory. I'm not holding my breath in the short run, kinda glad to have a job.
This isn't a new trend; it's just an acceleration in an existing arms race.
Automated HR screening has been a thing for decades now, as have mass application tools. The difference is just that instead of manually tailoring a resume for a job description (I am old enough to remember the trick of copy/pasting the JD and then minimizing & whiting out the font) the AI will do the automating for you.
Abstractly, the problem of resume review is of extracting a signal of candidate quality. That signal has become increasingly low quality. The various automated systems are a response to try and filter, but of course all filtering can be gamed and so the cycle continues.
Honestly at least in my industry everyone knows just blind applying for jobs is a pointless exercise; your probability of getting through the thousands of resumes that are going through the automated filter is very close to zero. Everything is about getting a high-quality referral. Which of course makes life hard for people who don't have a professional network!
Great read, Derek. Anecdotally - as an ancient Gen Z (27) it blows my mind to see the shift in experience from when I applied/got my job postgrad (5 years ago, how already?) to my friends searching for work now (1st/2nd/3rd jobs). AI, similar to COVID, feels like an inflection point of pre and post.
It brings to mind a thinker that I respect, Andy Crouch, who talks about technology as human's quest for magic (instant, effortless power), and how "Modern technology represents, in some ways, the most successful attempts we have yet seen of getting the world to do magic on our behalf."
But it comes with a cost on what it means to be human. Which raises the question for me- are we actually getting what we bargained for?
From my observation, whether it's online dating or applying to hundreds of job postings, we're increasing our pools/options but it feels like we're in a time of heightened scarcity…not abundance from the standpoint of human experience (relationships, face to face interactions both personally and professional, a sense of community, finding employment).