21–30 of 86 posts
"bone in a healthy animal will adapt to the loads under which it is placed.If loading on a particular bone increases, the bone will remodel itself over time to become stronger to resist that sort of loading. [...] The inverse is true as well: if the loading on a bone decreases, the bone will become less dense and weaker due to the lack of the stimulus required for continued remodeling."
Earlier quoted context omitted.
It'd be very fun and ironic if all this research, turmoil and energy waste just boiled down to a "better ad machine". I'm aware of other applications of AI and LLMs, but what most of the people see are the consumer facing ones like ChatGPT/Gemini/etc.
Well, what I don’t mind, are truly relevant ads. This is different from showing me (a man in my 60s) ED medicine ads, unsolicited, whenever I visit a news site. The behavior I need is typical “honest broker” behavior. These folks have been around for centuries. It’s a totally valid (and valuable) service. If I can present a specific workflow and requirements description to ChatGPT, and it responds with paid recommend…
As noted on the thread, a good ML model can do better, with less. I can use that kind of recommendation system for discovering my alternatives for a given item/tech/whatever, on demand.
But current crop of LLMs, with a strong leaning for embedding seamless ads into prompt responses, a double no. No, because I don't agree on how training data is stol ^H^H^H^H collected from wider internet and everything, another no for unsolicited ads.
I'll continue using Kagi and being a luddite, thanks.
[0]: https://research.google/blog/mechanism-design-for-large-lang...
One anecdote of this: in consulting, half of my brain power used to be spent on taking copious notes and organizing action items. Now tools like Granola have trivialized that process so i can focus on understanding
I’ve always said that we need to get out of our comfort zone, to grow. I’ve never thought it necessary to expound; it’s such a “no-brainer.” Right now, LLMs are selling a vision of “friction-free” productivity, but we haven’t yet gotten to the part, where humans start mastering LLMs as tools. That’s still quite nascent. Once that happens, a lot of people that have been firing all their employees, and getting fat on t…
Your prompts will be your moat and your margins will go to infinity. It will be glorious and sustainable for years to come.
- A large fraction of soon-to-be-redundant upper management, 2025.
Related: Wolff's Law "bone in a healthy animal will adapt to the loads under which it is placed.If loading on a particular bone increases, the bone will remodel itself over time to become stronger to resist that sort of loading. [...] The inverse is true as well: if the loading on a bone decreases, the bone will become less dense and weaker due to the lack of the stimulus required for continued remodeling." [0] https…
Earlier quoted context omitted.
Well, what I don’t mind, are truly relevant ads. This is different from showing me (a man in my 60s) ED medicine ads, unsolicited, whenever I visit a news site. The behavior I need is typical “honest broker” behavior. These folks have been around for centuries. It’s a totally valid (and valuable) service. If I can present a specific workflow and requirements description to ChatGPT, and it responds with paid recommend…
When I think about ads, I think about Google's paper [0], which allows multiple self-interested LLMs bid for possible ad-spots in the prompt, which is just a different flavor of what we have today. As noted on the thread, a good ML model can do better, with less. I can use that kind of recommendation system for discovering my alternatives for a given item/tech/whatever, on demand. But current crop of LLMs, with a str…
If LLMs can become true brokers, that would be great, but the pressure to corrupt would be tremendous.
Not sure if we’ll ever get there, though. I have hope. The broker model is centuries old. It’s really a solved problem. The issue is that the current generation is notoriously bad at learning from history.
Earlier quoted context omitted.
When I think about ads, I think about Google's paper [0], which allows multiple self-interested LLMs bid for possible ad-spots in the prompt, which is just a different flavor of what we have today. As noted on the thread, a good ML model can do better, with less. I can use that kind of recommendation system for discovering my alternatives for a given item/tech/whatever, on demand. But current crop of LLMs, with a str…
I guess I should clarify that I want a broker , not unsolicited ads. If LLMs can become true brokers, that would be great, but the pressure to corrupt would be tremendous. Not sure if we’ll ever get there, though. I have hope. The broker model is centuries old. It’s really a solved problem. The issue is that the current generation is notoriously bad at learning from history.
> The issue is that the current generation is notoriously bad at learning from history.
On top of that, the same people think they can just move to another planet and survive by eating money there.
Such a bold and non obvious statement that the author backs up with memorizing phone numbers as a youth?
Let me fix that:
> It is likely also going to be the biggest GENERATOR of creativity in my generation.
Reason: I am creative and I use it for creative things. And I no longer have to spend my memory holding phone numbers in my head.
Related: Wolff's Law "bone in a healthy animal will adapt to the loads under which it is placed.If loading on a particular bone increases, the bone will remodel itself over time to become stronger to resist that sort of loading. [...] The inverse is true as well: if the loading on a bone decreases, the bone will become less dense and weaker due to the lack of the stimulus required for continued remodeling." [0] https…
Same for muscles
Tools are only as powerful as their operators; and to be a powerful operator, you must face friction, the often the better.