What you see here is the last week’s worth of links and quips I have shared on LinkedIn, from Monday through Sunday.
For now I’ll post the notes as they appeared on LinkedIn, including hashtags and sentence fragments. Over time I might expand on these thoughts as they land here on my blog.
Product placement is a hit with consumers because it’s not as intrusive as an outright advert. Less so for studios and advertisers, because it’s a static arrangement. Just like a print ad.
Virtual Product Placement (VPP) aims to make this process more dynamic:
By using AI to insert brands/objects into TV scenes, you get the low-profile benefits of product placement along with the ability to change those placements on a whim. Having your cake and eating it, too.
If VPP takes off, in-show advertising could enjoy ad market liquidity similar to that of typical TV commercials and digital billboards.
That said … I wonder what this will do for TV show production? How soon till directors arrange scenes to leave more space for the (dynamic, digital, post-production) product placement?
Künstliche Intelligenz: ChatGPT befeuert Diskussion über Regeln
While reading these two articles:
1/ I keep thinking about the work that we try to hand off to LLMs, and the value we place on that work being performed by a human. (Especially when the human and the machine would produce roughly the same results.)
2/ I’m starting to see the “Prompt Engineer” role as an extension of AI itself: similar to the way an ML algorithm works backwards through a dataset to coax out the patterns therein, a prompt engineer tries to find the patterns within an LLM to make it usable.
(It’s like testing enough queries with a search engine to work backwards into the syntax: “put this term in quotes” and “put a minus sign in front of a word that should not be in the result.”)
Yesterday I put the finishing touches on the latest Data Science Salon Podcast episode.
I had a lot of fun talking with … Well, you’ll have to listen to find out!
Subscribe to the podcast and you’ll get the episode when it’s released:
Thanks to my interest in #risk , I’m constantly asking “what if?”, “what else?”, and “what next?” It’s all about getting a handle on what the future may bring.
Seeing what fiction writers produce is another way to consider those possible future scenarios:
(If you’re not familiar with PW Singer and August Cole: their books have been, to say the least, eye-opening treatises on the role technology may play in conflict.)
Say what you will about Facebook (or adtech in general) but it is interesting to see them trying a different approach.
Though, I suppose, external factors played a role. Losing $10B revenue due to Apple’s privacy practices would drive pretty much any company to change things up…
By taking charge of Advantage+ campaigns and using AI to test “different permutations” of ads on different demographics, instead of relying on advertisers to decide their targets, Meta is able “further those predictions” around what works best, according to Simon Poulton, vice-president of digital intelligence at digital marketing agency Wpromote.
Cody Plofker, chief marketing officer at Jones Road Beauty, said Meta’s new tool allowed brands to spend less time trying to work out how to wield Meta’s systems to target specific users and instead “focus on creative strategy” with ads that attracted more widespread attention.
The payments space is certainly an interesting one. Especially given the rise of pay-by-QR-code. (See: AliPay and WePay, in China.) And it’s hard to not see the parallels to cryptocurrency.
In particular, India’s UPI feels a bit like a central bank digital currency (CBDC). Not so much in the sense of “government-issued crypto tokens,” but as far as “nation-wide, government-supplied payment rails for digital transactions.”
I’m curious to see how this plays out longer term. What will UPI teach us about electronic payments? (And what lessons did they learn from WePay and AliPay?) A deployment this large is pretty much guaranteed to surface corner cases and other issues.
I’m glad that VCs are asking more questions about potential AI investments, but … they shouldn’t be the only ones performing this kind of due diligence.
“In the midst of the AI boom, investors are doubling down on diligence. Here are the five questions VCs are asking every generative AI startup during pitches.”
If you are:
- planning to buy a company that claims to do AI
- a CxO who is up at night wondering about their company’s AI efforts
- a prospective customer of some so-called “AI-powered” service
… then, just like a VC, you need to ask the tough questions.
“Is this company’s AI…”
- "… connected to the business mission?" (Not some exec’s vanity project)
- "… following industry best practices?" (Not cobbled together)
- "… living up to its potential?" (Not leaving money/opportunity on the table)
- "…prepared for the road ahead?" (Not sitting on a ton of unrecognized risk)
- "… real?" (Not smoke and mirrors)
And if you need help asking these kinds of questions … please reach out. I perform ML/AI assessments for just this reason.
Often when people talk to me about remote work, they simply say that “[some feature X] only exists in the office” and leave it at that.
And when I ask them “OK, so how would you recreate [X] in a remote workplace? What tools, techniques, or procedures would you need to make that happen?”
The usual reply? Silence.
Here’s a thought: instead of saying that remote work “can’t” happen, start asking what it would take to make it happen. Be open to possibilities. It’s much easier to find solutions once you acknowledge that a solution might actually exist.
So, hats off to Ethan Evans, below, for flat-out asking:
Leaders - please comment: what are the main challenges you face in sustaining high performance remote teams? I want to know so that we can develop solutions.
I encourage you to follow up on his thread and add your own solutions.