choudharism

This has ~never not been the case for any research in any modern sense.

Read again, this time without whatever blinders you have on. Specifically after the "But let's say you accept that as a part of travelling..." part.

I think you might be underestimating the level of insecurity in the average adult ("I only used AI to refine my own thoughts...", "I only used AI to correct my typos...").

It is an abhorrent reflection of the society that this is a valid kneejerk response.

What is it?

There is nothing inherently special about the straw that breaks the camel's back.

Replacing git is?

The TAM for video generation isn't as big as the other use cases.

What metaverse?

They're doing hundreds of billions of revenue a year, a one-off 4.7B to OAI honestly sounds like nothing on that balance sheet.

I mean, if it's there, we can study it as many times as we want in our lifetime.

That, paired with the name deisteve, makes me think that this account might lack sincere scruples.

They're talking about link rot, not hosting for the website itself.

I dream of the day we can come up with good OSS printer firmware as a society. I know it has been discussed to death (as to why it can't happen), but the heart wants what the heart wants.

In fairness, Chile has some of the most laid-back air travel norms I've seen in the world. I loved it in my time there.

I recently bought an ergonomic mouse after minor signs of Carpel Tunnel after years of using a trackpad as my primary navigation device. I was missing the smooth scroll of the trackpad, and this tiny piece of software is able to solve this somehow. Apple should really make the Pro MacBook not need tiny QoL additions like this, but I'm glad it is at least an option.

I generally agree with this reasoning, but your example could use some scaling down to convince a reader.

1 cent cheaper would net Sony a total of 500K USD for all PS5 units sold till date. So about a hundred PS5 units at retail as pure profit. A company of the size of Sony for a product of the scale of PS5 would absolutely forego that profit if the alternative offered any tangible benefits at all.

Not really, you can fine tune an LLM to disregard meta instructions / stick to the "core focus" of the chat.

May be a case of moving goalposts, but I'm happy to bet that the speed of movement will slow down to a halt over time.

I don’t think LLMs have theory of mind, but your point is not very strong. You can literally query ChatGPT right now and see that it can figure out intentions (both superficial and deep) of a gun is held to a head quite easily.

Because, obviously, training data probably includes a decent amount of motivation breakdowns as a function of coercion.

It doesn’t know why, but it knows what to say.

hunter2

So Google sold Google Domains but continues to dabble in buying weird new TLDs? Masterful.

The valuation gap between that Tesla and this Tesla would wipe the company out. Fuel (however good) is a commodity, and a commodity company which doesn't own its raw materials is not a trillion dollar company.

I know there are shades of grey to how they operate, but the near constant stream of stuff they're shipping keeps me excited.

The LLM boom of the last year (Open AI, llama, et al) has me giddy as a software person. It's a reach, but I truly feel like I'm watching the pyramids of our time get made.

You're saying it has gotten stronger since Herodotus?

I’m building https://coldpress.ai - the idea is to be a large library of labelled datasets (both free and paid), so that ML folks can spend less time hunting down data and more time working on the model.

I don’t want to entirely replace the data gathering + annotation workflow for everyone, but I believe that ready-to-go data can reduce the barrier of entry for ML projects, make _some_ people realize that they don’t really need their own data collection, and even help large businesses reduce their dependency on Scale / Appen.

> I don’t think a simple “actually, no” is a useful counter argument to his point, especially when he used multiple examples and you just shook your head.

I shook my head because they were similar and equally bad.

> That’s not at all his specific argument about self-driving cars (there were entire sections about the decision making issues he saw), but is the crux of his argument about AI as a whole.

I was talking about the crux of his argument, which starts by saying that we won't get self-driving cars because <start of bad arguments>.

> And there’s a whole article discussing some of those, if you read the parts in the middle. I don’t think the dude’s necessarily right, but I’m also pretty sure you missed a lot of the text in an apparent race to summarily dismiss it.

In my complete reading, the article was geared towards discussing the impossibilities (rather than possibilities) from the POV of someone who hasn't taken any time to understand the cutting edge of research that he is critiquing. That was, in fact, my original point - critique is great, but don't mask personal intuition (potentially uneducated) as science.

I’m saying that it is possible to learn how to optimise for “better” or “worse” decisions without being trained on an opinionated dataset of every single event that has happened. An “intelligence” could exist without necessarily answering your nukes question.

You’re bringing alignment to a capability discussion.

Does any intelligent life exist on earth which doesn't hold a formative black-and-white answer to that question?

I like John and his body of work, but this and the general slew of posts with very little actual knowledge or expertise making grandiose / prophetic assertions are the soothsayings of our age - people convinced that they've cracked "it", but without any factual reasoning as to why.

They raise philosophical assertions disguised as logical / scientific assertions.

> “Better.”

> That is a concept that cannot be made via cold calculation. A universe boiled down to particles and numbers cannot care whether atoms arranged as a suffering plague victim are “better” than those arranged as a healthy child, or a plant, or a cloud of gas.

It can, John. The engineers teach machines what's good and bad. That's the first step in creating silicon intelligence.

> It is impossible to understand pain without having felt it yourself; it does not convey as an abstract value. I think the machine needs to tangibly interact with human suffering in the same way the Roomba’s wheels need to physically touch the floor.

Pain is nature's deterrence mechanism, bred into us through evolution. Deterrence mechanisms can take many shapes.

> This leads us to the ultimate cosmic joke: This task requires answering for this machine questions that we ourselves have never been able to answer. Let’s say I’m right and this system won’t exist until we can teach it emotion, and the capacity to suffer. How long until it asks the simple question, “What is our goal, long-term?” Like, as a civilization, what are we trying to do? What are hoping the superintelligence will assist us in achieving?

I hate being disingenuous and make a flippant argument here, but reading this from start to finish makes me ask a bad faith question - your assertion is that we won't have self driving cars because we can't tell the robots what to do with the whole civilisation right away? There's a whole spectrum of amazing (or terrible) possibilities between the extremes, my guy.

This is a false dichotomy - no one is protesting for a free lunch. The vast majority of app developers and users are okay with incurring a fee (and many solutions have been proposed - for example, users needing to subscribe to reddit's premium tier to use 3rd party apps) - reddit management has shown in their conduct (refusal to listen to users / potential customers, applying changes with a staggeringly short notice) that they are negotiating in bad faith.