Upstream #202
beautiful drugs ??, stories ??, turing ??, reasoning ??, books ??.
Hey. how you doing? Winter is wintering isn’t it. Nice to feel the cold all the same. Seasons are special. Meaningful. Hope all good with you. Right, let’s go.
“The room is
almost all
elephant.
Almost none
of it isn’t.
Pretty much
solid elephant.
So there’s no
room to talk
about it”.
Kay Ryan (US poet and educator)
culture // beautiful drugs ??
We’ve always been appearance obsessed. Beauty standards, are an integral part of culture. But tech accelerates and distorts things. AI beauty filters for example, have worrying consequences, especially for teens and Zoom dysmorphia is a gateway to cosmetic surgery. We can get botox, lip filler, teeth veneers, or a facelift (even a low key ponytail facelift) with ease. Perm’agram face. You can even pay $12,000 to get your eye colour changed. They slit your cornea and inject your eye with dye. Jesus. And with the rapid growth of GLP-1 drugs like Ozepmic, weight loss is now effortless. Great for the obesity epidemic, less so for growing levels of appearance anxiety (“I’m not enough”). Ozempic may reshape industries, like snacks and high calorific products, but what will it all do to culture? Rex Woodbury wrote about how beauty is in the eye of the algorithm. How status, which is in part about cost signalling and scarcity, values that which has been difficult to acquire. Like Phd’s or Hermès bags (costly and scarce). So what happens when everyone is beautiful? Will that be valuable? Society is also aging rapidly, and these beauty standards will drive us insane. Just look at the ultra-bloody body horror The Substance to see this brought to its extreme conclusion. We shape our tools and then our tools shape us, and right now, they’re shaping us badly.
brands // stories ??
Three great stories to share. First, the amazing story of Rowntree’s (now Nestlé's) KitKat brand which began as a practical snack for 11’ses in work, and evolved to become the global symbol of break taking. It’s a great story about the power of strategic brand thinking, of iterating, learning and developing. It was nothing. It became everything. Michael Fitzsimmons from WeTransfer, wrote about building a brand people love and the qualities that have made WeTransfer so loved, for so long. How the brand feels and acts like an artist, embracing who they are for all of their faults, quirks, and mistakes and rarely talking about themselves. But importantly, how the brand isn’t built by the marketing team, it's out there in a million little pieces, and hands. In amongst the rules they live by, three resonated: actions over words, values over rules. coherency over consistency. And, amongst all the Jaguar noise, Volvo posted a 3 and half min story on Instagram. It’s great stuff, staying true to the brand's DNA of safety, but re-expressing it brilliantly. Here’s to the good stuff. Better to focus on that.
creativity // turing ??
This year Coke re-imagined their classic Christmas ad using AI. Eww. What? Yuk. But they did. And research company System 1 then tested Coke’s AI Christmas Ad with their large panel of very human beings, to see if it was well received and…wait for it…drum roll...it scored top marks. Wait. What? Top marks? You sure? Yes. Sure. Which begs more than a few q’s. System1 say, a lot of the love was built from the original “hand made” version. But still. Which marries beautifully with this piece on “how did you do on the AI art turing”. In a robust test of 50 pictures of either AI or human art, on 11,000 people, they found that most people had a hard time identifying AI art and wait for it…drum roll….most people slightly preferred the AI art, (even those who had said they hated AI art). Alan Turing’s test originally said that if 30% of humans couldn’t tell an AI from a human, then AI “passed”, so by these standards, it passes with room to spare; on average, 40% of humans mistook each AI picture for human. In other news Tesco champagne beats Mo?t in blind tasting and Charlie Chaplin came 20th in a Chaplin competition. We really don’t know what we think, do we.
technology // reasoning ??
In Sep, Open AI announced o1. They’ve merged the two key AI developments from the last 20 years: large language models (GPTs) and deep reinforcement learning systems (DeepMind’s Alphas). Things are evolving. But WTF does that mean? Well, we’re used the GPT model and LLM’s autofilled quick responses, the other is reinforcement learning systems, which is how Google’s AI powered Alpha Go beat the human Go world champion. Multi step reasoning machines, intended to solve hard problems. AI bloke Alberto Romero wrote a good piece on it, and the 10 implications for the future. He believes the best AIs will be reasoning-focused not generative and will go deeper, broader and faster than we ever could. He goes on to say "In the past, we were the architects and constructors of the world but then machinery happened, and we stopped being the constructors. We’re about to stop being the architects as well.” More reasoning models are launching too here and here. And when this tech improves and combines with other tech like robotics, things get interesting. It creates a revolution in how robots learn, which already involves very little explicit programming. The robots’ behaviour is learned with AI. Semi-autonomous humanoid housekeeping robots are coming, but it’s also accelerating life changing things like autonomous robotic surgery. Buckle up, things will get interesting and we don’t even know how.
Five random (ish) things:
watching // books ??
Books are nourishment for the soul ????. They break the shackles of time. They are proof we are capable of making magic. They are like seeds, that can lie dormant for centuries but may also produce flowers in the most unpromising soil. Right on Carl. Tóg go bog é.