whoisjuan

He might be right but this guy seems unhinged. Couldn't get past the feeling that I was reading the thoughts of someone who is unstable.

There are better and more articulate ways to make this argument. You don't need to hyperlink to another article at every phrase to build a sound, fact-checked and consistent idea.

Even before opening this thread, I knew people would be talking shit about Tailwind. I don’t get why HN seems to dislike it so much.

I’m just going to say that Tailwind, IMHO, is the best thing that has happened to CSS.

It’s by far the simplest and cleanest abstraction of CSS, and it accomplishes a lot through a very straightforward implementation.

At this point I can’t see myself writing CSS in any other way.

LLMs are to programmers what calculators are to mathematicians. I think LLMs will augment programming skills, not make them obsolete.

Care to explain why?

This is the kind of bureaucratic thinking that is resulting in Google being incapable of innovation.

“Run large language models like BLOOM-176B collaboratively — you load a small part of the model, then team up with people serving the other parts to run inference or fine-tuning.”

According to this excerpt, a node in the network doesn’t need load the entire model. Only a part.

I like the idea behind this because large AI seems to be highly constrained by co-located computation and the costs associated with it (GPUs and energy).

There are many delivery and cost advantages to running a massive LLM in a distributed P2P fashion.

Weirdly enough, I see this as a real "web 3" opportunity. Corporations running large LLMs could run their models on a decentralized network and pay participants for their contributed computing capacity.

AI most significant headwinds are cost and the pace at which GPU capacity is being built. This seems like a good model to tackle both issues.

They just keep putting out press releases talking about private betas and tests with "selected trusted partners" but nothing the regular consumer can play with.

If that's the pace at which they're planning to compete with Microsoft/OpenAI, they already lost.

People say that this SVB fail is isolated and idiosyncratic but I see a lot of stress in the system.

If there's no solution to this debacle by Monday this is going to keep cascading with no predictable trajectory of where it ends.

A simple Google search would show how much FB has contributed to the AI/ML community.

I hate seeing these comments in HN. No substance. Just saying whatever to jump into the FB hate bus.

I don't understand your argument.

So God can only exist if we are the only intelligent life in the universe?

That's some outdated geocentric reasoning.

"HN is more familiar with Coinbase's mistakes (not getting into the NFT scene earlier and letting OpenSea/MetaMask beat them to the punch..."

lol, who the heck thinks that not getting into NFT market was a mistake? If anything, they did themselves a favor by not building many NFT capabilities.

NFTs might have a second chance in the future but currently is just a mass graveyard of scams.

“best “

“bedt “

“old “

“wel “

“dta “

All of these crash safari in iOS for me.

What’s strange about this bug is that it happened overnight for multiple iOS version.

It seems to be a server bug that happens with the requests that populate the suggestions.

Turning off safari suggestions fixes it.

Groups and Marketplace are huge products. I presume they are larger or have become larger than the newsfeed.

The Bible is the fundamental book by which has driven many human accomplishments and debacles. You don’t need to be religious to understand it's impact in the modern human western society.

Still not as impressive as Vesna Vulovic, a flight attendant who holds the Guinness world record for surviving the highest fall without a parachute: 10,160 m (33,330 ft; 6.31 mi)

https://en.wikipedia.org/wiki/Vesna_Vulovi%C4%87

This will set a precedent though. Once they start taxing boosts they are just a couple moves away from taxing any kind of Ad buying.

What’s worrisome about this is that Apple is using it’s dominant position to possibly tax every transaction that happens in their ecosystem.

In a couple of years they could demand Amazon, Uber, DoorDash, etc. to pay fees too. Perhaps not 30% but they could certainly add a 3% fee to all those transactions. Something that looks reasonable to the public opinion and yet something that’s fundamentally fucked.

This is Apple operating like the IRS for their profit.

I have a couple of work relationships with people who don’t provide any verbal or physical cues that they’re listening and following what I’m saying.

It makes having conversations with those folks incredibly difficult.

It probably doesn’t help that we are having these conversations via Zoom, but I interact with other people via Zoom, and I don’t have this problem.

This is not pandering. I don’t have any information to categorically claim any of this, so why would I say it in any other way?

I’m self-deprecating because I’m on the fence on whether or not this is real or just some fictional plot I want to believe.

Chill, alright?

If you buy the theory that Bytedance is a covert operation by the Chinese government, then this fits the narrative.

Use the TikTok brand to attract top global talent and perform research in areas of general interest for government agencies.

They only need to build a backdoor for Chinese officials to access the research data.

Sounds far fetched but I believe this is well within the realm of possibilities.

You still need to prompt GPT-3 with something. Stacking two systems won't improve the outputs if your original input lacks precision. Garbage in, garbage out. Granted, GPT-3 requires way less precision to generate something useful, but you still need to pass the proper context.

I think you're right, though. It's hard to speculate how these things would turn out, but I think it's reasonable to assume that we will still need some relatively sophisticated human input to produce the best results out of these systems.

Learn how to write. It's as simple as that. These models will require well-crafted prompts to produce the desired outputs.

Your way to get ahead of this is to learn how the models transform a prompt into a video or image. You will quickly realize there's way more nuance and need for precise terminology than the shelf examples you're seeing.

These models are only as good as the prompts they receive. Someone needs to write those precise prompts. You are way ahead of everyone else because you already understand the concepts, terminology, and process of video production and CG.

This guy and his weird AI feud nobody cares about.

Why do people keep upvoting his stuff?

Ahh. Great catch. Their balloon definitely looks like a pin.

Unrelated, but why does the map on their homepage show a region in Cuba? That must be wrong.

So what causes heart disease? Trans-fats, carbohydrates, sugar, lack of fiber? ... I guess it has been evident to me for a while that saturated fat doesn't cause heart disease, but it's not apparent what does.

The problem with that isn’t the goal but the means to achieve that goal.

Perhaps if they trained their players with a collective mentality where the greater good justifies letting go of personal success, this would make sense.

But I’m pretty sure that’s not the case. Most Olympic athletes create a personal goal system, especially if competing in an individual discipline. Winning a gold medal has a profound meaning for them.

Helping their country to win a gold medal by giving up their personal chance to do it and winning a bronze medal instead has way less meaning.

So you’re not wrong, but it’s still pretty fucked up if you ask me.

"The SEC claims that the illegal trading by L. Palmer's former client and J. Palmer's friend netted approximately $35,000 and $73,000 in profits"... The SEC here going for the big whales. /s

Don’t sweat it. Thanks for the reply. And if it helps at all I agree that “binary” was the wrong word. I couldn’t come up with anything else that articulated what I was thinking.

What I meant is that us humans have this tendency to describe the world through dualistic models: good and bad, light and dark, mind and body, reason and emotion, nurture and nature, etc.

Inarguably this model has helped us to rationalize a lot of insights about the universe. However there are probabably a lot of ideas about our universe that can only be described through a dialectic model.

In fact, it was this type of thinking what allowed Einstein to image the paradox that gave birth to the general theory of relativity.

Look at the rest of the debate below the original answer, and maybe you will get an idea of what I meant.

My choice of words shouldn’t be the basis for attacking my points, especially since there’s more context in this very same thread that expands on my ideas.

Maybe instead of coming against me with an incendiary comment, make an effort to gain more insight into my views and add some value to the discussion. Again, my choice of words might be the wrong articulation of my ideas, but you just need to ask politely, and I will gladly expand and try to find better words.

Also, if you’re embarrassed to respond, then why did you? Please read the Hacker News guidelines if you forgot about them: https://news.ycombinator.com/newsguidelines.html