icapybara

I think we all want a viable Reddit alternative (and Twitter alternative), but the “fediverse” isn’t it. Why?

1. New user experience is very confusing. You have to learn a bunch of dorky sounding terminology right away (instance, fediverse/federated, etc), and there are a whole bunch of instances to choose from, but none of them may be particularly appealing or relevant to you specifically.

2. At least on Lemmy, I’m actively discouraged from signing up. And even if I ignore that, I have to fill out an application for an account that includes a personal statement on why I want to join Lemmy. Great, Lemmy can do whatever it wants, but it’s not a Reddit killer with a process like this.

This is a good counterpoint, but I would say that if they can’t make money off of having Reddit-like features, how are they making money off Discord in the first place? They aren’t that different. The reason to add Reddit features to Discord would be to kill a competitor (Reddit).

I know it doesn’t allow third party clients, but the difference here is that I like the discord app.

Discord should put out a quick Reddit clone. I imagine they already share a lot of the same users. How much would it take? Make the forum feature more like a subreddit for the server, improve discoverability of discord servers, make a “front page” that aggregates threads from popular servers...

This sounds kind of frustrating because it would be hard to predict or trigger. “Last time I ran into this error a super helpful window popped up, this time I’m getting nothing?”

Do you find this is a good solution for playing Steam games?

I have an older PC, and I'm thinking about replacing it with a macbook air and a steam deck. Does the steam deck feel more limiting than just having a windows PC with steam?

Interesting analogy, since the engine manufacturer also manufactures the car.

Anyone can build an OS without needing billions of dollars, and in fact there are state of the art open source OSes. Not quite the same.

The LLM _is_ the product. Everything else in the stack is window dressing.

I don’t think OpenAI’s contribution should be so understated- they built a technology that was considered science fiction just a few years ago. They deserve all the credit for the “AI”.

This point is understated. So many people are going around like "I'm building an AI app!" when the reality is, OpenAI built an AI app, you're just designing a front end for it.

Not many of them. Google or Microsoft can see what works for these startups and then put it in their own products. If you don't own the model, you don't have a safe business.

The article takes too long to get to the point and is wrapped in so many metaphors and newspeak that I feel like I'm being hustled.

It is possible to write about this topic clearly and concisely.

> There's an extra constraint involved by requiring that i^2 = -1, which could be written other ways, that ties the two elements of the tuple together. It seems quantum physics requires this constraint in order to describe reality.

I think you couldn't have put it better. I'm just a physicist, though.

"When you hear hoofprints, think horses, not zebras."

You can always contrive a more complicated structure to explain some observation. There is no evidence branes exist. We are looking for an explanation to the "dark matter problem" that doesn't require asserting that a ton of new and unproven physics exists.

Thanks ChatGPT

Those are software applications, but AI is going into hardware as fast as it can. Imagine a turret gun that can see and think, and makes decisions about targets. Or a robot/drone that uses AI to navigate its environment.

Correctness is also a concern. You can be sure that a program you write will do that, but you can't be sure that the LLM will do that correctly every time.

For many companies this type of promise is not useful. It doesn’t matter that they say they won’t, they still can look if they want to. This is the primary concern when you’re dealing with trade secrets where the secrecy of the information is its only protection.

Others have answered your question, but I'll add that the market for high quality AI models is not similar to the software marketplace, where there is always an open source alternative (and where open source is often the state of the art).

LLMs take so much engineering effort, research, and compute that it's unlikely there will be good open source alternatives in the near future. Right now your only real option is OpenAI (or maybe Anthropic) and that seems unlikely to change anytime soon.

The only reason we have LLAMA is because Meta threw us a bone. They might not do that again.

You are way over simplifying.

It is not remotely intellectually challenging to go to the moon. It just takes rocket fuel. Newton solved motion hundreds of years ago, and now high schoolers compute it in physics class.

> there will likely be plenty of alternatives soon to fall back on

Are there? OpenAI is currently the only provider. Big tech companies (Google, MS, Meta, maybe Apple) may also soon have similar models but they don't have much incentive to sell access to them. Smaller orgs can't produce GPT-level AI. It feels like all these products are really at the mercy of OpenAI.

There has been an explosion of effort around LLama, but Meta did the hard part (the base model), and who knows if they will do that again.

It's interesting to see all these companies building products wholly off of OpenAI's platform. What happens if OpenAI changes their mind about pricing or access some day? Isn't this an existential risk to all "AI" companies that don't actually do the AI part?

You hope. Personally, I've seen Bing ask me questions about my prompts that were surprising.

I’m not arguing they stuck with their original position. I’m saying the name is still reasonable. I don’t care if they changed positions.

An open API you can build products off of. Consider the alternative. They could not let anyone use their models and just release AI products (their own search engine, their own version of Teams, their own self driving car software, their own photoshop competitor, etc).

Should they not be able to make money off a world changing new technology they bet billions on?

Does your company offer all its products for free?

Well just the other day people were shocked at how cheap GPT3.5 was.

Why would it know?

Some machines might last longer than us but they are still subject to the same laws of thermodynamics. Most computers from 20+ years ago are inoperable today; show me a machine that can survive millions of years.

It's physics defying. Optical resolution is directly correlated with the lens diameter, there's no way around this.

Images are formed by interference of plane waves propagating at different angles. To resolve finer details, your optical system needs to collect waves that were propagating away from the source (the moon) at larger angles. This means you can either get closer to the moon or increase the lens diameter. It doesn't matter how the lens works, you could use any photonics magic trick you want, you're still limited by the (aperture) size of the optical system.

Samsung's lens is purportedly too small to resolve the level of detail produced in the moon images.