What We’re Reading: The struggle to understand A.I.
Hi everybody,
A few weeks ago, we shared with you our latest thinking on A.I. generated writing on Medium, how we’re working to keep it out of your feed as well as out of the Partner Program. It’s clear that A.I. is a permanent part of our digital lives, for good and ill (Meta rolled out a new chatbot to Instagram and WhatsApp this week). When we hear from readers and writers on Medium about A.I., it’s a mix of curiosity and apprehension.
Luckily, there are stories on Medium to speak to both sides. This extremely long read from Meta data scientist Colin Fraser goes deep on why A.I. hallucinates, and gives a great grounding in machine learning along the way. And this in-depth look at what Apple might be up to with the future of Siri — it involves lighter LLMs and interface recognition — hints at some exciting new possible ways A.I. will make doing certain tasks easier.
On the flip side, the relentless pace of A.I. development makes having the bigger — and much needed — discussions about its impact across society and government hard, but one writer argues it’s essential we not cede the conversation to the vested interests behind big A.I. companies. And if you’re worried about the A.I. future and need any clearer signal that your anxiety is shared, browse the stories that show up on Medium when you search for “AI replace.”
What have you read recently that’s changed or deepened your understanding of A.I.? Hit respond and let us know.
In the meantime, thank you, as always, for reading.
– Scott Lamb, VP of Content at Medium
What We’re Reading
The Customer Is Always Wrong
Published by Avi Siegel in Entrepreneurship Handbook
Customers are so rarely right, especially about SaaS products. And even more, especially when they graciously (if not heatedly) tell you their ideas about how the product should or shouldn’t work or what it should or shouldn’t do, whether big or small.
The Matrix: 25 Years On
Published by Simon Dillon in Fanfare
Anyone with a penchant for anagrams will immediately have the answer spoiled for them. Still, what’s fun about The Matrix isn’t so much what happens but how it happens. The exhilarating final forty-five minutes in particular are an action-packed blast, as Neo finds himself up against the notorious Agent Smith (Hugo Weaving); surely one of the most sinister and menacing sci-fi villains of all time. As self-aware computer programme nemeses go, Smith is a superbly nasty piece of work, and his battles with Neo are edge-of-the-seat stuff.
- A nuclear fusion reactor in South Korea set a new record for sustaining a temperature of 100 million degrees Celsius for nearly a minute earlier this month. It’s a small but important step towards fusion becoming a viable energy source.
- “How I Discovered the Secret to Giving Feedback in the Workplace” is story from a former CEO, sharing years of hard-won wisdom on feedback.
Today’s Final Word goes to writer and editor Chris Anderson, who’s been regularly updating his “Nerd Urban Dictionary” since 2020. The post, as Anderson describes it, is “a crowdsourced list of sci/tech/biz words that nerds use to signal that they’re smart. These are perfectly good words in their native sci/tech/biz context, but risk being pretentious when used u̶t̶i̶l̶i̶z̶e̶d̶ in day-to-day conversation.”
A few examples:
- “Orthogonal” instead of unrelated (“My political views are orthogonal to this conversation about you finishing your salad, young lady”)
- “Non-trivial” instead of hard (“Shipping the code by the end of the day is non-trivial, boss”)
- “Ping” instead of “remind” or “check up on”
- “Bandwidth” instead of capacity (“I don’t have enough bandwidth for that”)
- “Signal-to-noise” instead of just quality. (“The signal-to-noise of the comments on that post was super low”)
Read more — including probably some words you use (I’m personally guilty of all the examples above!) — here.
Read or write anything new? Finished that draft from Medium Draft Day? Share it in the responses!