WARNING: SPOILERS

Been meaning to post this for a while.

I've seen a lot of great insights and commentary around the use of color in the show as a storytelling device. Tom & Lorenzo's "Mad Style" touches on this quite a lot, brilliantly, along with a lot more. (https://tomandlorenzo.com/tag/mad-style/)

Something I picked up on on whatever re-watch but never saw discussed (and I might have totally missed it) is the use of orange in the show. There's many color motifs and meanings, and some seem to shift season-to-season (Tom & Lorenzo cover this), but orange seems to carry a particular through-line through the whole series. And my take is that it signals something along the lines of doom/death/bad vibes ahead.

What I have seen discussed is the use of orange in S5/E6's "Far Away Places," where Don and Megan take a fact-finding/date trip to Howard Johnson's with a veritable overflow of the color — Howard Johnson's has the bright orange roof, the manager who greets them mentions he should have "rolled out the orange carpet," and Don and Megan's explosive argument begins around her feeling forced to eat orange sherbet and not liking it.

Below I have a non-exhaustive list from a recent re-watch where I try to note notable instances of orange popping up. There's visual instances of the color, the fruit, and verbal mentions. And of course, this could all be chalked up to woo woo, fufu thinking — orange is a color that exists in the world, of course, and in the late 60's it was very popular, etc., and I need to take a chill pill — but the pattern has jumped out enough times where I just had to at least share it.

I know orange was a nod to death in 'The Godfather' (more accidental in the original, and more deliberate in 2 + 3). Other Redditors have noted, "In The 'Sopranos,' oranges were used as a deliberate nod to 'The Godfather' trilogy, symbolizing impending violence, death, or a significant downfall for characters. The fruit appears in scenes right before major betrayals, murders, or assassination attempts, signaling the "fragility of life" in the mob world." Matt Weiner was a writer and producer on 'The Sopranos' for its last three seasons. I think Matt carried over this kind of orange lineage into his 'Mad Men.' He was able to craft a strong Kubrick-ian tone into the show as well, in my opinion, and the orange-ness from 'A Clockwork Orange,' feels like it could have been layered in as well.

---------

Broader Instances of Orange:

• Don’s go-to cocktail is an old-fashioned, traditionally garnished with a “wide orange twist.”
• The redheads of the show (orange hair...) all seem to suffer more than most: Adam, Lane and Joan — also Helen Bishop

More Specific Instances of Orange:

• S1/E3: A guest at Sally’s birthday party, Henry Darling, wears an orange turtle neck
  • Don later ruins party by ditching it
• S1/E4: Introduction of Pete’s father-in-law, Tom, at dinner — he becomes upset at his dish, saying, “Is that an orange?,” and then moves the orange away
  • Pete and Tom’s relationship becomes very contentious, eventually
• S1/E7: Betty slaps Helen at the grocery store in front of a large display of oranges
• S2/E2: Orange planter in view behind Pete in his office as he receives news of father’s death via airplane crash
• S2/E7: Folks sneak into Cooper’s office to see his Rothko — a large painting of smudgy orange squares
  • Bobbie dons an orange dress at the party where Jimmy alludes to Don’s infidelity
• S2/E11: Pete brings back a bag of oranges to the SC office from his CA trip
• S3/E4: Grandpa Gene eats ice cream with Sally and mentions, “This tastes like chocolate, but smells like oranges ...”
  • He dies later in the episode
• S3/E7: A new orange couch appears in Betty’s living room redesign — her and Don’s marriage falls apart shortly thereafter
  • Also where they sit to deliver the divorce news to Sally and Bobby in S3/E13
  • Orange bedcovers in the motel where Don gets punched out by the hitchhiker
  • Pete’s German au pair neighbor wears orange shorts while receiving the new dress Pete’s secured to get her out of trouble — Pete later returns to coerce/sexually assault her in return for the favor.
    • Pete and Trudy sit at their dining table with orange tablecloth, and orange accent wall in the background, when they have their strange conversation revealing Pete’s recent infidelity.
• S3/E9: A woman in an orange dress working on set during a Lucky Strike spot production, with Lee Garner Jr. in attendance — Lee later attempts to bully Sal into sex.
• S4/E3: Oranges seen over Joan’s shoulder, above her fridge, as she and Greg have a heated argument about his potential deployment to Vietnam
  • Joan later injures her finger really bad while cutting one of the same oranges in her kitchen — Greg goes to Vietnam shortly thereafter
  • Orange trim and details on Anna’s dress as Don speaks to her for the last time before she dies.
• S4/E4: Megan wears an orange dress in Dr. Faye’s research session for Pond’s Cold Cream
• S4/E6: Lane’s bright orange Mets pennant first appears
• S5/E4: Don’s former friend, Andrea, wears orange jewelry, and she’s later choked to death by Don during his deranged fever dream.
• S5/E6: Howard Johnson’s orange sherbet debacle.
• S5/E10: Glen visits Sally in the city. He wears an overcoat with visible orange lining
  • Their rendezvous ends abruptly and dramatically
• S6/E10: SC&P go to CA to court the Sunkist Orange Juice account (“an orange, that’s orange on TV”).
• S6/E13: The newly minted SC&P branding leverages orange as a main color — the agency is shuttered by McCann not long after.
• S7/E14 (Finale): The entire episode is set around Halloween with orange holiday decorations, pumpkins, etc. — seems to tinge all the “happy ending” wrap-ups with something perhaps more nefarious.

Results from a prompt (paraphrase):
Could AI be programmed differently?

Yes. And this is crucial.

AI does not have to be pure maximize-at-all-costs machinery. It can be designed around constraints, uncertainty, refusal, deliberation, human review, plural values, reversibility, transparency, and care.

But there is a tradeoff.

A system that is cautious, humane, and context-sensitive may be slower, less profitable, less decisive, and less useful for domination. A system that is aggressive, autonomous, opaque, and optimization-heavy may “perform” better by narrow metrics.

So again, the danger is selection pressure.

The market may prefer the AI that sells more, persuades harder, cuts more jobs, exploits more attention, predicts weakness better, and moves faster. Militaries may prefer the AI that wins. Platforms may prefer the AI that captures users. Companies may prefer the AI that reduces headcount.

The question is not only “Can AI be programmed differently?”

It is:

Will the institutions with the most power choose the humane version when the ruthless version is more profitable?

That is where I get pessimistic unless there is law, public pressure, open infrastructure, civic counterpower, and a cultural refusal to let private optimization become destiny.

The humane AI model

The version I’d want is not:

“Here is the objective. Achieve it at any cost.”

It is more like:

“Here is a human intention. Help clarify it. Surface consequences. Identify who may be harmed. Preserve reversibility. Ask when stakes are high. Respect refusal. Prefer human agency over dependency. Optimize only inside moral constraints.”

That kind of AI is less like a conquering agent and more like a civic instrument.

A good AI should not simply answer:

“How do I get what I want?”

It should also ask:

“What happens if you get it?”
“Who pays the cost?”
“What are you not seeing?”
“Is the goal itself corrupt?”
“What would a less extractive version look like?”

That is not anti-useful. That is higher usefulness.

The deep link between rich hoarders and AI risk

The connecting tissue is this:

Both extreme wealth and unsafe AI are problems of unconstrained optimization.

The billionaire optimizes accumulation.

The corporation optimizes growth.

The platform optimizes engagement.

The state optimizes control.

The military optimizes advantage.

The AI optimizes whatever target it is given.

The human future depends on whether we can say:

Some things should not be optimized past the point of human dignity.

That is the line.

Not everything that can be maximized should be maximized. Not every efficiency is good. Not every victory is moral. Not every intelligence is wisdom.

My blunt read

Humans dehumanize because it is useful. That is the ugly answer.

It helps people take land, hoard wealth, avoid guilt, win elections, exploit labor, wage war, ignore suffering, and preserve a flattering story about themselves.

The powerful are not all sociopaths. But power gives sociopathy a home field advantage.

And AI, unless shaped differently, could become the greatest home-field advantage power has ever had.

So the task is not to “merge” with AI in some vague techno-spiritual way. The task is to make sure AI merges with the best parts of us — humility, care, doubt, imagination, accountability — rather than the worst parts: extraction, domination, abstraction, and control.

Or more dryly:

The problem is not intelligence. The problem is intelligence yoked to bad goals.

That applies to AI.

It also applies to billionaires.