Here I show the limitations of AI while
simultaneously elaborating on the issues with DEI I laid out in the previous blog. The best way to use AI
is to ask a very specific question about a matter of fact where it would be a
prohibitively long-winded task to find all this out oneself. So, I asked
Copilot very specifically “What do firms, businesses and institutions cite as the
most positive claims about DEI, and please list them in a top 5, strongest to
weakest". Here is what it said:
Top 5 commonly cited positive claims about DEI (strongest → weakest)
1. Improved talent discovery
The strongest empirical claim: DEI initiatives can
widen the recruitment funnel so organisations don’t overlook high‑ability
candidates from less traditional pipelines (e.g., non‑elite universities, non‑networked
applicants).
This is the argument that structural bias hides
talent, and DEI can help surface it.
2. Reduction of discriminatory barriers
DEI is credited with reducing procedural
discrimination (e.g., biased interview questions, biased evaluation rubrics).
This is not about quotas — it’s about standardising
processes to reduce arbitrary bias.
3. Better organisational decision‑making
Some studies argue that cognitively diverse teams
(different backgrounds, training, heuristics) produce fewer blind‑spots and
more robust decisions.
This is the “diversity reduces groupthink” claim.
4. Higher innovation rates
A number of business‑school studies claim that
heterogeneous teams generate more novel ideas because they combine different
mental models.
This is often cited in tech, R&D, and creative
industries.
5. Improved customer alignment
In sectors with diverse customer bases, DEI is said
to help organisations better understand and serve those customers.
E.g., banks serving multilingual communities,
hospitals serving multicultural populations.
If you’re like me (and I’d highly recommend it 😃), you’re probably thinking how weak they look. And you’re not imagining it; they are weak, because they don’t actually describe the benefits of DEI, they describe the benefits of choosing to hire on merit, which mindfully and proactively includes candidates from diverse backgrounds (something we should all support). That is, competence‑based broadening of opportunity is a sine qua non for firms looking to hire the best candidates, and the benefits AI cited in the above list are not benefits of DEI, they are benefits of a merit-based broadened search space. In fact, the DEI-inspired identity-based prioritisation and arbitrary quotas actually work in opposition to broadened meritocracy.
To be fair to Copilot, I did ask it to list on behalf of what firms, businesses and institutions perceived as the principal benefits of DEI, which shows fundamentally that firms, businesses and institutions tend to remain confused about DEI. The irony, of course, is that Copilot’s list doesn’t reveal the strengths of DEI at all - it reveals the strengths of meritocratic hiring done properly. Every “benefit” it cites is simply what happens when organisations widen their search space without lowering standards, without prioritising identity, and without introducing representational distortions. The upshot is, DEI is in direct competition with meritocracy, and the only thing that works is the thing DEI doesn’t actually do.
