contra Brian Balfour on LLM Platforms

https://blog.brianbalfour.com/p/the-next-great-distribution-shift 🔥 my verdict^°hard word: it stinks^°hard word!

There are several glaring^°hard word factual^°hard word or interpretation^°hard word errors that cloud his analysis^°hard word of the space.

----

But first, his analysis^°hard word is that "tech^°hard word platforms^°hard word" go through a three-stage^°hard word cycle^°hard word:

• Identify^°hard word the moat^°hard word
• Open the gates
• Close for monetization^°hard word

His analysis^°hard word of how this worked for multiple^°hard word large tech^°hard word companies, unfortunately^°hard word, is fatally^°hard word flawed^°hard word.

----

Facebook - to start with, several of the "added features^°hard word" like Marketplace^°hard word and Photos existed years before the ill-fated^°hard word "Facebook^°hard word Platform^°hard word" ever^°hard word launched^°hard word. But, the real problem is that Zynga-style^°hard word gamification^°hard word cluttering^°hard word activity^°hard word feeds was a disaster^°hard word 💡 although, in hindsight^°hard word, it was probably better than the political takes that followed it. They could not allow it forever, so it stopped. And, without a limitless^°hard word reservoir^°hard word of free advertising^°hard word, Zynga^°hard word crumbled^°hard word.

Also, constantly^°hard word throughout^°hard word this post, the author^°hard word is nostalgic^°hard word for what can best be described as awful^°hard word crap^°hard word. The "quizzes^°hard word" and "vampire^°hard word bite" apps were not something to be defended^°hard word; the people who ostensibly^°hard word "made millions^°hard word" from them not to be celebrated^°hard word.

----

Apple - first off, the "70-30" revenue^°hard word distribution is considered as both "open" and "closed". Once again, people being able to make millions^°hard word off the "iFart^°hard word" app is not to be celebrated^°hard word. And the demands^°hard word that the platform^°hard word never include^°hard word features^°hard word that are in platform^°hard word apps are unreasonable^°hard word.

But the more serious concern^°hard word is the specter^°hard word of government regulation^°hard word. Most of the "privacy^°hard word restrictions^°hard word" he criticizes^°hard word are imposed^°hard word by the government. And the anti-trust^°hard word concerns^°hard word govern^°hard word pricing^°hard word more than anything^°hard word else.

----

Google - the SEO^°hard word ⚙️ search^°hard word engine optimization^°hard word industry is largely the scourge^°hard word of the earth, and should be destroyed. 💡 there seems to be a trend^°hard word that it is bottom-feeders^°hard word, creators^°hard word of frivolities^°hard word, and abusers^°hard word of the public common that suffer from this "contraction^°hard word". Unlike^°hard word the author^°hard word, overall^°hard word I would prefer those changes

The problem here is that the moat^°hard word is contrived^°hard word and unrelated^°hard word to the changes. The increase in oneboxes^°hard word ⚙️ in Google^°hard word parlance^°hard word, a "onebox^°hard word" is a structured-content^°hard word response^°hard word, other than organic^°hard word search^°hard word or an ad^°hard word. For example, a "weather" onebox^°hard word for a search^°hard word for the weather. They were called "oneboxes^°hard word" because there would be only one per^°hard word search^°hard word result page, but that seems to have gone by the wayside^°hard word. was certainly a change from the 2004 philosophy ... but it's unrelated^°hard word to market-share^°hard word growth against Yahoo^°hard word.

----

LinkedIn - I'm not even sure what he's talking about. I don't believe there was a "2 year window" where LinkedIn^°hard word encouraged^°hard word content^°hard word creators^°hard word. I think it was either a COVID^°hard word effect, or a personal^°hard word observation of this guy^°hard word extrapolated^°hard word to the whole^°hard word site^°hard word.

And, once again, the people creating "B2B Marketing" content^°hard word are people I do not want to succeed.

----

So the conclusion^°hard word, that platform^°hard word cycles^°hard word are accelerating^°hard word, cannot be supported by this data. And the attempt to project this onto ChatGPT^°hard word is bad.

The moat^°hard word has to be chat^°hard word memory ... because this guy^°hard word can't think of anything^°hard word else. And, even if it is ... the platform^°hard word that can access^°hard word your email history is more valuable^°hard word than the chat^°hard word history.

Also, ChatGPT^°hard word can't shut down the API^°hard word. For one, it's profitable^°hard word. For two, the alternative^°hard word of "self-hosted^°hard word models" is too good for this to meaningless^°hard word harm^°hard word the (now)-competitors^°hard word.

The ChatGPT^°hard word "market share" advantage^°hard word is driven^°hard word by higher name recognition^°hard word among non-technical^°hard word people, and people who haven't tried competitors^°hard word. They are in the position of Yahoo^°hard word! and MySpace^°hard word. I'm not saying that this means they can't be successful^°hard word ... but there is not such an advantage^°hard word that Google^°hard word and Facebook^°hard word are irrelevant^°hard word.

And, finally, the complaints^°hard word that ChatGPT^°hard word is competing^°hard word with enterprise^°hard word players^°hard word like Glean^°hard word are pathetic^°hard word. Presumably^°hard word the feature^°hard word is "connectors^°hard word" ⚙️ https://techcrunch.com/2025/03/17/openai-to-start-testing-chatgpt-connectors-for-google-drive-and-slack/ : being able to access^°hard word data in your Google^°hard word Drive account. This is a feature^°hard word that Claude^°hard word has had for a long time. Once again, this is a ludicrous^°hard word demand^°hard word that a platform^°hard word remain incomplete^°hard word because someone else implemented^°hard word a necessary feature^°hard word.