Desire, dopamine, and the internet ⊗ Search engines for personal AI agents ⊗ A learning network

No.304 — Everything is interesting ⊗ City scanning ⊗ Nvidia 2024 AI Event ⊗ Speaking without vocal cords

Desire, dopamine, and the internet ⊗ Search engines for personal AI agents ⊗ A learning network
Ferry seats by Josh Hild on Unsplash.

Desire, dopamine, and the internet

You might remember a few weeks back, an article by Ted Gioia that made the rounds quite a bit, The State of the Culture, 2024. My one phrase summary was that, in his thinking, “art was consumed/subsumed by entertainment, which was then overtaken by distraction, and currently by addiction through algorithms and dopamine hits.”

The reliably smart and thorough L. M. Sacasas let it simmer quite a bit more and argues in turn that “the dopamine framing, while grappling with real and important dynamics, is inadequate and may ultimately be counter-productive.” As per his usual M.O., he then digs through a number of thinkers and ideas to understand what the deeper and longer term dynamics have been.

Sacasas considers what’s missing in the central graphic, looks at addiction vs compulsion, and the writings of Blaise Pascal who argued “that human beings will naturally seek distractions rather than confront their own thoughts in moments of solitude and quiet because those thoughts will eventually lead them to consider unpleasant matters such as their own mortality, the vanity of their endeavors, and the general frailty of the human condition.” It leads L.M. to our current endless diversions, to Philip Rieff’s anti-culture, and coping with super-abundance.

I encourage you to take your time and read both. Side note: eight years after everyone else, I’m currently reading (about half-way in) Cal Newport’s Deep Work. In chapter three he explains why deep work is meaningful and gives neurological, psychological, and philosophical arguments for depth. The whole thing is angled for deep work of course, but I have a feeling they can be read also as hints of what people are missing when feeling anxious, depressed, silent quitting, etc. And I feel like that ‘thing’ he’s pointing to overlaps quite a bit with Sacasas’ argument and with why we fall prey to Gioia’s dopamine culture. Just a couple of feelings, maybe I’ll get back to it, hit reply or the comments (link at the bottom of this email) if it resonates or if you want to explain why I’m wrong.

Today we live under conditions of information super-abundance, but we sometimes still operate under the presumption of information scarcity. For example, we are tempted to ravenously consume information as if we might somehow miss out on something of value or else find that the information has somehow run out before we’ve had a chance to take some for ourselves. […]

This new anti-culture, he explained, “aims merely at an eternal interim ethic of release from the inherited controls.” Permissions all the way down. Which, it is absolutely worth noting, happens to correlate remarkably well with the demands of a consumer economy. […]

In a culture of information superabundance, we need above all else the discipline to say “no” or to set limits upon our engagement with the vast proliferation of digital media. But the anti-cultural spirit has left us ill-prepared to say “no” to anything. […]

This is the question we need to be asking and attempting to answer, for ourselves and for others. We need a compelling account of silence, solitude, attention, disciplined engagement, well-considered restraint, vulnerability, and risk. But not for their own sake or for the sake of nebulously resisting the lure of digital technologies, and much less out of a misguided reactionary impulse. Rather, we must come to see these as the necessary skills and requisite virtues for the pursuit of our well-being and that of our neighbors.

Who will build new search engines for new personal AI agents?

Another favourite writer, in a completely different register, Matt Webb looks into what’s needed for personal AI agents to arrive, what’s been delaying them, and how they might work. As much as I like his reasoning, I had a constant vibe of ‘yes, but why?’ Kind of like when looking at the ai pin, rabbit r1, or The 01 Project (at least this one is open source). Cool tech, but frivolous use cases and societally irrelevant. Not as in ‘they will have no impact,’ but as in ‘they will have none of the impacts we need.’ Good piece by Matt, likely quite an accurate take, I just have a hard time getting exited about these things, considering the energy impact of AI—even if he’s talking about edge usage, not centralised like ChatGPT.

Any answer regarding the future of AI agents must also answer the “there from here” question. I refuse to believe in a near-term future where AI agents somehow displace my iPhone, or require me to have another device in my pocket. […]

We’ll achieve this in the exact same way that we meet the challenge with phones – user preferences are basically “vibe” and this is expressed as “brand” for an “app” in the phone world. The skeuomorph of that app squircle isn’t going away, that’s my take, the “there from here” challenge is insurmountable otherwise.

Not a social or professional network, a learning network

I tweaked the title of this piece by Tom Critchlow so you wouldn’t skip over the name LinkedIn. But yes, it’s about how he proposes that they should change how they view the company and become a “personal learning network for every user—centered on their personal development.”

When Microsoft bought LinkedIn, my take was that they wanted a direct connection to people so they could sell their tools and services to individuals, in a world where so many of us change jobs often and fluctuate between employed, employer, and freelancer—they always did of course, but sold a lot more through licences for corporations, and stores. Tom is roughly in the same direction but in a much much much more elaborate and intriguing way. Even if you don’t care about that company, still a good read for a vision of something that is needed, whether it’s through them or someone else. (Also, 100% on the highlighted quote below.)

The world of work needs radically new models for how we learn and what we learn. Not just because of AI, but that ain’t gonna help for sure. We need new models of soft skills, interpersonal skills, creative skills and faster deployment of technical skills to keep up. […]

The magic here is that LinkedIn is already the best place for learning. I took the time to re-orient my feed from old connections to people I actually wanted to follow and now my LinkedIn feed is higher signal to noise than Twitter, Threads, Bluesky and the rest of em.


§ How to write. The honest truth? Everything is interesting. Absolutely everything. Sewage systems, the psychology of UFO sightings, the effect of school testing, the behaviour of cells. If it's approached correctly, every single aspect of life is potentially fascinating. But once your curiosity is in play, it's crucial to pay careful attention to its opposite. You have to notice when you're getting bored.”

Futures, Fictions & Fabulations

2024 AXA Foresight Report
Good/unique looking report. “These stories show some of the extraordinary advances we can expect when different fields come together, be it science, technology, society, nature and art, powered by a collective confidence to build a more positive vision of the future. Women and young people will emerge as agents of change. Businesses also have a unique opportunity to make a difference in areas such as climate change, diversity, and mind health.”

Future Days - City Scanning
“This was a new thing I had not tried before, but combining the different Focus Areas of the Future Days initiative, based on the four scanning layers (loosely based on Richard Buchanan’s Four Orders of Design) seemed useful. It gave people a quick map across which they could check what signals they were seeing, what else they might relate to, and just generally gave a prompt to consider more than just what was it front of them.”

2024 Digital Media Trends
“This year’s Digital Media Trends point to continued industry disruption—not just from streaming, social, and gaming, but also from how these media and technologies combine. As these disruptions mature, their convergence sets up the next revolutions. Companies need to lead these advancements—or risk living in another’s vision.”

Algorithms, Automation, Augmentation

Nvidia 2024 AI Event: Everything revealed in 16 minutes
“Nvidia CEO Jensen Huang kicks off its GTC keynote in San Jose with a slew of AI infused chip announcements.” Lots and lots and lots of acronyms, loads of products. Worth a watch, even if some parts might be too technical for some, just to see the breath of AI products they are making and how many fields they are sliding into.

“The king is dead”—Claude 3 surpasses GPT-4 on Chatbot Arena
“On Tuesday, Anthropic's Claude 3 Opus large language model (LLM) surpassed OpenAI's GPT-4 (which powers ChatGPT) for the first time on Chatbot Arena, a popular crowdsourced leaderboard used by AI researchers to gauge the relative capabilities of AI language models.”

Sakana AI’s evolutionary algorithm
“The core idea of Sakana AI is to use natural principles such as evolution and collective intelligence to create new AI models. The goal is to develop a machine that automatically generates customized AI models for user-defined application domains, rather than developing new models each time.”

Asides

Your Futures Thinking Observatory