garrickvanburen

I agree. We don't need a nondeterministic 10quadrillion vector model. We need an deterministic expert on our narrow business. Something small, that can be run on the 2026 version of the spare PC under the CTOs desk.

I think Dan Olson and Folding Ideas is doing a fantastic job of bringing thoughtful criticism to all kinds of modern media, most recently Mr. Beast Games.

https://www.youtube.com/channel/UCyNtlmLB73-7gtlBz00XOQQ

Agreed. Also, as every media organization is free to make editorial decisions on both what they cover and how they cover it, left/right is often far too simplistic and vague to actually reverse engineer a media orgs bias.

I agree.

Between: more efficient models - tuned for the task at hand, the ability to run those models in-house, or even at the edges, plus Google and Microsoft are well positioned to stay ambivalent as they’ve got lots of products to sell and whether or not LLMs are part of the portfolio mix is completely dependent on enterprise customer demand.

Anthropic/OpenAI have a number of aggressive downward pressures on their pricing.

it's a great reminder that Uber wasn't a 2-sided marketplace to begin with, just an on-demand black car service, and Travis drove early on. The marketplace model came later, copying Lyft, more as a low-cost expansion strategy than a business model.

This. Stated another way, you need to start by either: fulfilling existing demand yourself....or being the demand yourself.

My default position: If an LLM can create it, we probably don't actually need it.

I agree, and a willingness to experience failure as soon as possible has always been a competitive advantage.

If anything LLM chatbots & synthetic users will make the majority of founders evermore comfortable not testing reality.

https://dejabru.org - The Homebrew Competition Remembering Historic and Long-Forgotten Beer Styles

I flip this around.

Marketing comes first. Sales second. Product third.

“The business enterprise has two—and only two—basic functions: marketing and innovation.” - Peter Drucker

I'm pretty sure my primary job is marketing the work that I do.

Considering: - Anthropic decides how much a token is worth. - Users have no visibility or ability to control in how many tokens a given response will burn.

This is the only expected answer. https://forstarters.substack.com/p/for-starters-59-on-credit...

Based on my math, taking development in-house doesn’t make sense until we’re talking >$5K/mn.

https://forstarters.substack.com/p/for-starters-72-vibe-code...

I'm reminded of the original Agile joke, "software you don't want in 30days or less." today it's "software you don't want in 5days or less."

The most hyped use cases for AI/LLM make me wonder, "why are we doing this activity to begin with? We could just not."

If I recall correctly, the impetus for Christiansen to even look into this was to answer the question: “why are there so many milkshakes sold in the morning via drive thru?”

That’s a different question than, “how do we make our breakfast menu more popular?”

Now, perhaps the latter question begat the former, but either way it is not a product innovation story, it’s a customer/market research story articulating the shape of customer demand (which I fully support).

If there was a second part of the story - like, McD started marketing milkshakes in their breakfast menu and sales shot up ##% or they developed a breakfast flavor milkshake and sales shot up - then we’d have the product innovation story.

I'm a big fan of JobsToBeDone, but partially because of this milkshake example, it took me years to grok.

As I see it, the milkshake example isn't about innovation, it's about actually having human-to-human conversations with customers and appreciating customers define their own competitive set, so perhaps closer to Gibson's "the street has its own use for things."

Outside of those two points, I'm not sure what repeatable lessons exist in the milkshake story for innovation, and because of that, it's probably doing more harm than help to JTBD as a concept.

I would say Apple's original exclusivity deal w/ AT&T shows they're comfortable letting a 3rd party control some of the user experience if that's the only way they can get to market.

It’s likely the land would be far more valuable as something else.

Small family farms, while romanticized, have all the problems of any small business competing with larger professionalized businesses; consistency in operations, consistency in output quality, access to resources - including people and machines.

Additionally, for their own operational simplicity big buyers prefer interacting with as few suppliers as possible - so, market forces have been driving consolidation for decades.

On the news of Netflix acquiring Warner Bros, I’m reminded of how good Netflix has been at innovating their business model.

Over the past 27 years, their business model has changed multiple times and each evolution appears to be in direct response to the bottleneck of growth, from maintaining inventory of DVD to acquiring global streaming rights.

Year / Business Model / Bottleneck to Growth

1998 / Sell DVDs over the internet / Need to continually replenish DVD inventory,

1999-2006 / Rent DVDs over the internet / USPS delivery & return times

2007 / Stream movies over the internet / Acquiring US streaming rights to a massive library of movies

2009 / Start producing movies (Netflix Originals) / Number of subscribers watching Netflix Originals

2010-2012 / Global expansion; Canada, South America, Europe / Maintaining rights globally

2025 / Acquire Warner Bros Discovery

I prefer Gell-Mann Amnesia Effect or Knoll’s Law of Media Accuracy

“AI is amazing about the thing I know nothing about…but it’s absolute garbage at the stuff I’m expert in.”

https://garrickvanburen.com/an-increasingly-worse-response/

My rant on token-based pricing is primarily based on the difficulty in consistently forecasting spend.....and also that the ongoing value of a token is controlled by the vendor...."the house always wins"

https://forstarters.substack.com/p/for-starters-59-on-credit...

Fortitude is amazing. Cinematography, soundtrack, creepiness all top notch.

"social media" is forums, IRC, blogs, etc, but through the lens of advertisers.

One of my favorite phones. Despite a bizarre hiccup preventing sd cards from being recognized if the camera lens sensor broke.

After briefly looking into it, my assumption is intermittent fasting works great for people that are eating throughout the entirety of the day.

Or Kubb! It has a high silliness factor