Jump to content

Artificial Intelligence

Parsad

By Parsad
in General Discussion

 Share

Recommended Posts

  • Replies 823
  • Created
  • Last Reply

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

Popular Posts

Peregrine
Peregrine

Because these are real problems afflicting the country that wouldn't cost even a fraction of what AI spend is costing now. More importantly, no one even knows what the hell this AI spend is going to b

bizaro86
bizaro86

I think it's possible shopify is an AI beneficiary. I signed up for a Shopify store today, and was having trouble getting the DNS settings at my domain registrar to match what shopify needed to host t

Posted Images

Valuebo Community Regular

Valuebo

Posted
Posted

SA: "I don’t want to quibble on the exact definition of the Turing Test. By the popular conception, we kind of have passed it and most people in 2020 did not think that was going to happen. So now if in the next five years we really deliver on AI, there’s discovery and important new science and we were hyping our progress and our excitement a little bit too much, I think we deserve some grace there, but it’s annoying and we should do less."

 

---

 

Read: we made an advanced chatbot that can hardly add up numbers so we'll definitely get AGI for sure in 5 years time. 

 

Dude keeps bullshitting over subscriptions, integrating ad models and adding third party apps as if all those things would matter in a few years if his little plaything actually ever made even half an actual scientific discovery in the next few years, let alone achieve true AGI. Moving the goal posts nonstop.

Spekulatius Grand Master

Spekulatius

Posted
Posted (edited)
2 hours ago, Valuebo said:

 

So basically integrate third party apps' chatbots, search functions etc and call it AI? 

 

Brilliant! I'm sure we'll get our AGI any day now!

I think it’s a little more complicated than that. The vision seems that they hand off tasks to different apps, then move them along workflows in ERP environments. Basically, what I think companies like UI Path tried to do, but easier to implement.

 

Or maybe the vision is just to replace Google, but that would be more for consumers. Given how well Gemini/ search works, I think the latter is very very unlikely. 

Edited by Spekulatius
Eldad Mentor

Eldad

Posted
Posted

All of the conmen coming out and saying their false promises are annoying and we are in a bubble while at the same time accelerating the madness daily, is very weird.  

Castanza Grand Master

Castanza

Posted
Posted
2 hours ago, Spekulatius said:

Sounds a bit like OpenAI wants to be an operating system of sorts:

https://openai.com/index/introducing-apps-in-chatgpt/


The Motley Fool actually had a good discussion on this topic today. Only like 20 minutes but they raise a lot of concerns and honestly I agree. “A lot of AI seems like a solution looking for a problem.”
 

How is it monetized?


How do the partners work?


Which partners get priority and why?

 

Is it sharing my data and asks with every other partner?
 
How does this “widen the funnel” for customer acquisition? 
 

Examples given: 

 

User says: “I want to get in shape.” 
Does said user just get spammed with ads from the highest paying partner? 
 

User said “I want to fly to Minnesota.” Which partner gets the traffic? Booking? Or Expedia? Who determines that? 
 

Overall I think they call out a lot of the utopian promises. At the end of the day I think people will prefer decision making over better curated spam advertising. 


https://podcasts.apple.com/us/podcast/motley-fool-money/id306106212?i=1000730880886

 

 

Xerxes Grand Master

Xerxes

Posted
Posted

Bezos in Italy with John Elkann as the host. On technology and AI 

 
full interview. 

 

The 10 min on AI bubble where Bezos schools Elkann 

 

 

 

Libs Proficient

Libs

Posted
Posted
6 hours ago, MungerWunger said:

https://www.uncoveralpha.com/p/too-much-ai-too-soon

<What also doesn’t get enough attention is that much of the spending by current tech leaders is not tied to new revenue streams but actually to defend the moats and business models they already have. They are in a race that has gotten out of hand, but as Meta’s Zuckerberg has recently stated, the risk of overspending a few hundred billion on infrastructure is smaller than the risk of being left out. >

 

Good article. Thanks. This is the money quote ^^^^^^.

 

Thanks to the late great Charlie Munger, I live by the rule, "show me the incentive, and I'll show you the behavior."

 

Zuck's quote is so telling. When this whole thing collapses, it will become famous. Of course Meta, GOOG, MSFT, will be just fine when the bubble pops. But there will be  a ton of smoldering wreckage by the side of the road.

 

Oh, and in 15 years the world will be transformed by AI. Just like railroads, electricity, the internet. It just feels so obvious to me that AI will follow the same path. 

 

Pelagic Rising Star

Pelagic

Posted
Posted

AI's water use gets brought up a lot by its detractors. I thought this article's attempt at comparing query water use to everyday items was interesting. Sort of tangential but I've always found how homes in Bermuda are designed to capture and store nearly every drop of rainfall that falls on them interesting, who knows we might see data centers adopt something similar just so they don't have to deal with municipalities claiming they're causing local water issues, whether real or perceived. 

 

 

 

Hektor Mentor

Hektor

Posted
Posted

Is this the beginning of a super app? I wonder what this means for the META apps family. 

 

https://www.wsj.com/business/retail/chatgpt-walmart-shopping-3e411e83

 

Soon You’ll Be Able to Shop Walmart in ChatGPT. Here’s Why It Matters.

Retail giant signals that online shopping is about to change

 

Walmart is forming a partnership with OpenAI to let shoppers buy its products directly within ChatGPT, the artificial-intelligence chatbot. It is a signal by the biggest U.S. retailer that online shopping is going to become a totally different experience from the retail websites we are all used to.

 

Within the next few months, U.S.-based ChatGPT users will be able to instantly buy Walmart products directly in ChatGPT. The products will include nearly everything available on Walmart’s website, except for fresh food. And Walmart+ members will still get their benefits such as free shipping when making purchases through ChatGPT.

 

ChatGPT already allows shoppers to buy some products from Etsy’s domestic sellers and will soon add some merchants on Shopify’s e-commerce platform. 

bizaro86 Mentor

bizaro86

Posted
Posted
On 10/3/2025 at 9:07 AM, billybobjovialdechicoutimi said:

Maybe it's been discussed before... pardon me if so and thanks in advance for a link to the relevant thread... but

 

How is it that the market is terrified that CSU's business will be crushed by AI, but nobody has such fears for Shopify... (or at least, its stock only gets an erotic caress)

 

Can AI not reproduce in one second Shopify's entire infrastructure?

 

Thanks for any insights, I am not smart enough to get it myself

 

I think it's possible shopify is an AI beneficiary. I signed up for a Shopify store today, and was having trouble getting the DNS settings at my domain registrar to match what shopify needed to host the site. I asked their AI assistant, which looked up my existing DNS settings and told me exactly what to change and then it worked. I was really impressed, it made it super easy.

  • Like 1
rogermunibond Mentor

rogermunibond

Posted
Posted

Dwarkesh Patel

@dwarkesh_sp

·

Oct 17

The most interesting part for me is where

@karpathy

describes why LLMs aren't able to learn like humans. As you would expect, he comes up with a wonderfully evocative phrase to describe RL: “sucking supervision bits through a straw.” A single end reward gets broadcast across every token in a successful trajectory, upweighting even wrong or irrelevant turns that lead to the right answer.

 

> “Humans don't use reinforcement learning, as I've said before. I think they do something different. Reinforcement learning is a lot worse than the average person thinks. Reinforcement learning is terrible. It just so happens that everything that we had before is much worse.” So what do humans do instead?

 

> “The book I’m reading is a set of prompts for me to do synthetic data generation. It's by manipulating that information that you actually gain that knowledge. We have no equivalent of that with LLMs; they don't really do that.”

 

> “I'd love to see during pretraining some kind of a stage where the model thinks through the material and tries to reconcile it with what it already knows. There's no equivalent of any of this. This is all research.” Why can’t we just add this training to LLMs today?

 

> “There are very subtle, hard to understand reasons why it's not trivial. If I just give synthetic generation of the model thinking about a book, you look at it and you're like, 'This looks great. Why can't I train on it?' You could try, but the model will actually get much worse if you continue trying.”

 

> “Say we have a chapter of a book and I ask an LLM to think about it. It will give you something that looks very reasonable. But if I ask it 10 times, you'll notice that all of them are the same.”

 

> “You're not getting the richness and the diversity and the entropy from these models as you would get from humans. How do you get synthetic data generation to work despite the collapse and while maintaining the entropy? It is a research problem.” How do humans get around model collapse?

 

> “These analogies are surprisingly good. Humans collapse during the course of their lives. Children haven't overfit yet. They will say stuff that will shock you. Because they're not yet collapsed. But we [adults] are collapsed. We end up revisiting the same thoughts, we end up saying more and more of the same stuff, the learning rates go down, the collapse continues to get worse, and then everything deteriorates.” In fact, there’s an interesting paper arguing that dreaming evolved to assist generalization, and resist overfitting to daily learning - look up The Overfitted Brain by

@erikphoel

 

. I asked Karpathy: Isn’t it interesting that humans learn best at a part of their lives (childhood) whose actual details they completely forget, adults still learn really well but have terrible memory about the particulars of the things they read or watch, and LLMs can memorize arbitrary details about text that no human could but are currently pretty bad at generalization?

 

> “[Fallible human memory] is a feature, not a bug, because it forces you to only learn the generalizable components. LLMs are distracted by all the memory that they have of the pre-trained documents. That's why when I talk about the cognitive core, I actually want to remove the memory. I'd love to have them have less memory so that they have to look things up and they only maintain the algorithms for thought, and the idea of an experiment, and all this cognitive glue for acting.”

Eldad Mentor

Eldad

Posted
Posted

The fake videos are getting crazy good.
 

Funny that the court room standard for evidence is probably about to return to the ancient standard of 2 eye witnesses as video is about to become complete garbage. 

Castanza Grand Master

Castanza

Posted
Posted (edited)

I have been listening to Sam Harris's Podcast recently and worked through some of his AI discussions. The conversation on ep #434 was interesting. The book written by the guests Eliezer Yudkowsky and Nate Soares "If Anyone Builds it, Everyone Dies" is pretty interesting read thus far.

 

Also working through the podcast series suggested by Sam Harris "The Last Invention"...Not going to lie, it seems a bit out there for a suggestion from Sam since he is typically a very rational and clear thinker. Frankly I'm not sure what to make of it....but only into ep 2 so perhaps it comes full circle...

 

The more I listen to and read things like the above, the more convinced I am that this whole capex stuff does not matter. This tech is going to be built Hell or high water. 

 

Edited by Castanza
rogermunibond Mentor

rogermunibond

Posted
Posted
31 minutes ago, Dalal.Holdings said:

No worries, I'm sure it will all work out

 

 

 

 

 

Not all capex is created equally.  How much of that is for electric, site, build and how much of that is GPUs, servers, cooling, back up power, etc.

 

What happens to the useful life of the H200 GPU cluster in five years?

 

 

 

Milu Experienced

Milu

Posted
Posted
51 minutes ago, Dalal.Holdings said:

No worries, I'm sure it will all work out

 

 

 

 

I would suspect that many of these companies revenues and earnings are up 10 to 20x over the last decade too, so not sure how much signal you can get from these charts.

Dalal.Holdings Grand Master

Dalal.Holdings

Posted
Posted
5 minutes ago, Milu said:

I would suspect that many of these companies revenues and earnings are up 10 to 20x over the last decade too, so not sure how much signal you can get from these charts.

 

The only thing that matters is the returns generated on that capital.

 

I'm highly skeptical that it works out to decent returns, let alone the massive ROIC software and ads have generated in the past for these firms.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share
Go to topic listing


×
×
  • Create New...