Friday, November 22, 2024

Ad Review: Samsung "The Next Big Thing Is You"

 Lately I've been feeling my marketing roots. I also somehow managed to set some setting on my phone which can't use "Ublock" ad-blocker, to not personalise ads. I've never been impressed by targeted ads, and I mean how could I be, I basically buy nothing. 

Anyway the first one I'm going to review is fairly simple, but interesting, I would argue because the struggle is real, the struggle being of course to sell benefits of "AI" more precisely LLMs, which we've had for years.

Here was an ad that kept rolling during the Paris Olympics:


Though by my recollection the "Can I kick it?" was translated into a Paris themed French. What is curious is that speech-translation apps have been available since 2010, certainly this benefit has been available on smart phones since 2015, so LLMs are being sold on the same benefit that pretty much everybody already has. 

It should be stated that said benefit I can go for years without seeing anybody use it. The last time I saw someone use a translation app was in 2018 at a hotel I was checked into in Mexico City, a pricey one in which to sleep off my jet lag, an Indian business man was using an app to ask for more bottled water and more towels of the receptionist. It was translating English to Spanish, and I could speak to the receptionist in English (My Spanish would remain shit for years). 

This benefit is in itself, if not a dud, very very marginal. Though it might seem impressive that the whole world is now open up to you. If you think about it, the major benefit of this app is that it would allow me to take a two week trip to Hungary without having to learn any Hungarian, especially not a lot of Hungarian. 

Beyond this, if I was to go live in Hungary for a year though, do I want to be using an app to talk to people? To ask for bacon at the deli? In a multicultural city like Melbourne you just never see this. Furthermore unlike countries outside the Anglo-sphere where increasingly pretty much everyone in the services industry needs to speak English anyway, relatively few Australians speak a second language. 

So here we have arrived at scratching our heads in respective marketing departments to sell LLMs. I pick on this ad, but the overwhelming marketing message of LLMs is selling it as "AI" this mysterious thing that is possibly an existential threat that we need to have in order to stay ahead of it.

The fact of the matter being, that Google Translate from 2010, maybe 2004, is as much "AI" as the current flood of ambiguous LLM based "AI" products. Chat-gpts ability to produce paragraph length coherent responses to plain English questions is an impressive if not necessarily useful leap forward in software. 

Also please note in the extended version of the Ad, the kid who upon reflection is clearly British first asks a group of Spanish speaking women "can I kick it?" to which they respond "Que?" (What?) and then he pulls out his phone AND PRETRANSLATES EARLY AFROCENTRIC RAP LINGO INTO PLAIN ENGLISH "can I play?" before the phone displays "puedo jugar?" on the outer side of the flip - which is probably the real feature they are trying to sell, and the group of girls responds "si, puedes." (yes you can.) Now if the AI could do a LLM translation that went and analysed A Tribe Called Quest lyrics to understand the sentiment that would truly be impressive if unlikely given that the phone would first have to deduce the context in which the user is speaking at all, and even if the app can access the phone camera and microphone and is like "oh Edward is on a basketball court with some girls...analyzing..." The intelligent conclusion to reach is that what he means is "can I kick the basketball with my foot" being that the user is British and they play football with their feet.

"Puffery" is as a legal term in reference to an advertisers legal ability to exaggerate the benefits of their product or service. So it is kind of interesting that in the full version of the Samsung Galaxy ad, they actively avoid puffing up the abilities of LLMs, as in that they can understand Q-tip and Phife Dawg (RIP) employed in a context where not only is the slang obscure and dated for the native speaker but he is choosing to employ acute slang when communicating with a foreign culture.

Bringing me to another dud product sold on the buzzword in the zeitgeist:


Though not timely, I must confess its the Lumosity ads I really want to review. Because they really spoke to my sensitivities of marketing education. More so than the "can I kick it?" Galaxy ad that features heavily a white suburban US youth, its hard to say who that ad is for. 

Lumosity ads were targeted at morons. They used to be all over youtube, I would see lumosity ads maybe two or three times a day. The most striking thing was that all the spokesmodels for lumosity come across as dumb. 

Compare it to an ad for weight loss, or male pattern baldness. In these ads, shonky as they are, what one would expect is a before/after comparison. Here I was really fat, and thanks to Nutronic meal replacement system here I am now thin and athletic. Here I was bald, and thanks to Nutronic hair replacement system here I am now, full head of hair and confident again.

The logical ad structure for lumosity's brain training is showing someone who was dumb and is now smart. Instead we just got ads populated by dumb people talking about how they heard the game was based on "neuroplasticity" which everyone is talking about.

The ad takes the predatory tact of praying on insecure people's need to be relevant. And I suspect most advertising that features "AI" is taking the same tact here in 2024. Albeit circumstances are somewhat different:

For one thing, Lumosity was the really prevalent player spruiking a product with the buzzword "neuroplasticity", it may be a failing of my memory but I can't remember any other brain-training apps or company advertising. But pretty much everyone has AI tools, likely thanks to OpenAI being both the leaders in developing LLMs and being quite open about how they were doing it.

Nobody really "owns" LLMs, the closest would be OpenAI with ChatGPT which I would bet like $20 has the highest brand recognition. Midjourney certainly was leading in AI art for years, but I can name five other AI art brands where for most product categories I would struggle to name more than 2. 

And unlike neuroplasticity which was more or less a successful rebranding of the concept of "practice" LLMs are useful tools. They've arrived, and not only arrived but have been here for years in lesser powerful versions like Google search. I doubt LLMs are a mere fad.

But that's where the meaningful differences stop. The rest is marketing white-bread to the white-bread crowd. I mean who wants a camera that takes photos that swap out your children's actual expressions for happy ones? Parents that want to send their kids to North Korea, that's who. 

Perhaps I've neglected a meaningful difference, and I don't want to write a post on "AI" so much as review an ad that perplexingly can't sell "AI" as being any different to a product we have had for 10 years already. 

Lumosity got sued by the FTC and had to stop selling their snake oil on its curative properties:

On January 5, 2016, Lumos Labs agreed to a $50 million settlement (reduced to $2 million subject to financial verification) to the Federal Trade Commission over claims of false advertising for their product. The Commission found that Lumosity's marketing "preyed on consumers' fears about age-related cognitive decline, suggesting their games could stave off memory loss, dementia, and even Alzheimer's disease", without providing any scientific evidence to back its claims. The company was ordered not to make any claims that its products can "[improve] performance in school, at work, or in athletics" or "[delay or protect] against age-related decline in memory or other cognitive function, including mild cognitive impairment, dementia, or Alzheimer's disease", or "[reduce] cognitive impairment caused by health conditions, including Turner syndrome, post-traumatic stress disorder (PTSD), attention deficit hyperactivity disorder (ADHD), traumatic brain injury (TBI), stroke, or side effects of chemotherapy", without "competent and reliable scientific evidence".[10][11][12]

There is no good medical evidence to support claims that memory training helps people improve cognitive functioning.[13][14] ~ from Wikipedia AKA basically 90% of the Gemini LLM generated Google search responses.

This may be why products and services that now sell AI in my experience do not talk about the supposed benefits of AI. They don't even do the "gotta be the shoes" thing from the old Nike Spike and Mike ads where there would be legal recourse to claiming no reasonable person would actually think it is Michael Jordan's shoes that make him the greatest basketball player of all time. (now the richest on the other hand...)

Curiously gambling ads in Australia are steering well clear of claims about "AI" pushing instead social media and "Actual intelligence" but ads for daytrading software are pushing AI, and I suspect that's because while advertisers have figured out daytrading is for all but the tiny population of professionals working for large financial institutions or for extremely rich private clients, mathematically identical to gambling - in so far as if 70% of users lose all the money they put in.

Probably the best ads at really trying to sell a benefit, rather than a feature of LLMs, is the one I've seen a few times but cannot remember what it was for - that kind of said that with AI tools maybe you could do the job that used to require expertise - like a fashion designer or car designer. It is still however a struggle to sell this as a benefit without being predatory - specifically the only people who will be sold on this point are likely to be morons. Why? Because if you can suddenly design a hypercar or a pair of hightop sneakers with no particular expertise SO CAN EVERYONE so there's no benefit there to your career. Same as uber income being driven down by all the people who can become drivers for uber.

So in conclusion, the only scumbags using AI to prey upon dumb people are the trading company ads and there probably are some gambling app ads that do claim AI can figure out the ponies for you. That's really bad in the moral/ethical sense. Companies like Samsung and Apple and Google and Microsoft I view more as struggling to come up with ways in which LLMs are useful. 

Don't get me wrong, they are useful. Just it seems only marginally more useful than the products and services we already have. Hence search engines are trying to force us to adopt Gemini output now, and Microsoft downloads an AI companion when you update windows even though you never used Cortana apart from maybe asking Cortana how to uninstall Cortana. 

Video calls were a hard sell at first, until basically people got used to doing them on desktop computers and phones switched from minutes to data plans. I'm sure some of the LLM features will prove as beneficial as video calls did. At the moment I am underwhelmed though, I don't think marketers at least have cracked the code of selling LLM features by their benefits.

No comments: