Since I've seen no one talk about this, I'm coining the phrase 'Middle Child Phenomenon'.

A law student who entered university four (4) years ago is faced with a curriculum that became completely outdated two (2) years in. Let's take a cohort of 1000 law students from 2020, and explain from there.

• 100 students drop out by 2022 for a myriad of reasons; sickness, family, poverty, academic difficulty, etc.
• 80%/720 students are the nobodies who cruise by, haven't really become a lifer yet.

What happens to the remaining 180 students?

• 50%/90 students are the all-rounders, committed enough to do well academically, have time for extracurriculars and do an apprentice at a Big 5 law firm.
• The other 50%/90 students are split into two (2) groups: (i) the CV-Chasers, these win competitions, and can recite the constitutional precedents of the last decade, and (ii) the Golden Children who do everything, and they do it damn well.

The Golden Children of around forty-five (45) students talk here and there about artificial intelligence, maybe host a few workshops. They point out a giant, GIANT, GIANT problem with alignment, and it goes something like:

Richard: "Hey, why doesn't our curriculum have any artificial intelligence electives? Why aren't there any professors including homework problems, or real-life examples for us to deal with?" 

Sally: "As they say: '...born too late to sell textbooks, born too early to fix the justice system but born just in time to read the gajillion discovery files Saxmon just gave me. Let's-s-s-s go-o-o--'

Jake: "Hang on for sec, this is an actual problem. Either we beg our supervisors to let us work with Dr Bellows on that first AI case he's cooking up, or we have to do our own reading."

Richard: "And if we don't do something now--since these Professors who've never entered a courtroom in their lives--we're going to graduate in a completely different world."

Sally: "Then get replaced by the young hotshots who are blessed with a curriculum that actually gives a damn...I'm getting it now."

Richard: "It's a shitfest in the shitfactory on planet shitstorm indeed."

The point is, forty-five (45) students is 4.5% of every yearly cohort. Even less than that will be good at helping the war effort with AI alignment, and that's assuming they get hired in a firm with the raw ammunition necessary to institute motion applications that materially change legislation and get parliamentarians to do something.

My conclusion: 

• Make bridging courses ASAP. Presently, medico-legal courses include doctors, and medical practitioners to familiarise law students with the field in bite-sized pieces.
• Programmers and alignment researchers need to get on board with forming expert committees that include the younger generation. 
• Create a platform like Brilliant.org that's designed to be simple and inviting for law students and other related fields.
• Workshops and expo's that introduce law students to artificial intelligence in an exciting, and breathtaking way. The kind that gets a fire going, a sparkle in their eyes and makes them say: "Yes, this is what I want to specialise in."
• Start shifting the narrative away from: "LLM's are destroying education!!!" to "Let's integrate basic programming skills, and work problems (without corroding critical thinking) into the average law student's life."

If you don't, you'll get the 'Middle Child Phenomenon'. A generation of lawyers who, for a period of three (3) - five (5) years of lawyers, either have no interest in AI, or worse, have the wrong ideas on how to go about it. 

The long-term consequences of ignoring this sect can be catastrophic. Europe's AI Act is a clear example. terribly written from a legal standpoint and even worse when it comes to alignment efforts. 

It's time to stop ignoring the middle child in the family, shovel some food onto their plate and see what strength they can muster. I'll be writing a post on what forms that strength can take, and elaborate more on some initiatives to fix the problem, soon. For now, I wanted to alert the community.