šĀ On time for your weekend: a round-up ofĀ this weekās remarkable storiesĀ at the intersection of technology, business, design, and culture. Three reads and three listens; no fluff, just stuff ā”
There is absolutely no inevitability as long as there is a willingness to contemplate what is happeningāMarshall McLuhan
š Reading
Faith and Fate: Transformers as fuzzy pattern matchers:
The larger implication here is this:Ā rather than think of models as working through the parts of a problem in a general and systematic way, it is often more accurate to think of them as search engines, recalling examples which roughly match that specific part of a problem before, and then stitching together such approximate recollections.
Alexis GallagherāAnswer.AI | 15 minutes
Seeing like a CEO:
Complex human institutions are their own realities, whose most important characteristics subsist in habits and relationships and invisibly distributed skills. Effective leaders certainly do strive to surface and measure objective, quantitative data about their institutions, and incorporate it into their decision making. But those kinds of measures cannot be remotely sufficient to guide and govern the living organism that we call a firm.
Steve Randy WaldmanāInterfluidity | 8 minutes
Monetization & Monopolies: How The Internet You Loved Died
āTech Monopolies: They profit by creating positive externalities for society. They donāt last that long. And their rents are both reasonable and self-defeating.ā Progress itself depends on the natural movements of this cycle being allowed to run unchecked, despite the unease many feel when looking at these companies in isolation [ā¦] I understand the temptation to seek alternate solutions. But doing so is worse for consumers, worse for other founders, and worse for society as a whole.
Conrad Bastable | 40 minutes
š§ Listening
Rot Economics:
Generative AI is actually an informational tool. So you should use generative AI as a way of generating, filtering, summarizing, finding, checking information. That's actually what it's good at [ā¦] I think many people are going to use it for automation because that's theĀ why. That's what companies are being told [ā¦] It's the thing that they may experience from other technologies and it's the thing that some people are telling them that that's what they should do.
Daron AcemogluāWhereās Your Ed At? | 33 minutes
Zuckerbergās Open Source AI Manifesto
There's a whole story about open versus closed; distributed power distributed risk versus centralized risk. It is funny that this week we had the Zuckerberg letter and Llama3 talking about let's make sure we decentralize AI a little bit and open source it in a way that you can run different things and we have like this great example of what happens when things get over-centralized in terms of risk factors.
Sam LessināMore or Less | 57 minutes
Into The Metaverse:
Whether you're talking about artificial reality or virtual reality, both of which come from the late eighties in science fiction, or cyberspace [ā¦] that was all based then on what people believed we were soon going to do. And now 30 years later, those terms are being repurposed, readopted, becoming to some extent part of corporate propaganda terminology [ā¦] All of them were efforts to vaguely describe something that was starting to manifest.
Matthew BallāInfinite Loops | 89 minutes
š Timeless
1ļøā£ year agoāWhy havenāt internet creators become superstars?
2ļøā£ years agoāThe future of tech as I see it
3ļøā£ years agoāA Spreadsheet Way of Knowledge