an0malous

Wasn’t that YC startup Delve doing exactly this?

https://www.iansresearch.com/resources/all-blogs/post/securi...

-

I said this in another thread a while ago, and one of these people who thinks Apple isn’t on the side of privacy cited a lawsuit they settled around Siri listened to conversations: https://www.scientificamerican.com/article/apple-settles-cla...

People understood this settlement to mean Apple was spying on their conversations and selling them to advertisers, when it seems to have more to do with people accidentally triggering Siri. But people don’t care about this kind of nuance or actually tallying up all the ways Apple is pro privacy against rare issues like this one. It’s all just tribalism at the end of the day.

Why? It’s a position of less security and stability

So he thinks open weight models will lead to “AI communism” and “dystopian hell” and in the very next point proposes that the US create a federal agency to discourage the use of open weight Chinese models. The motivated reasoning in this post is unreal.

This would be awesome if someone could add 3D rendering to make it actually look like I’m pressing a little jelly ball. It would make it so fun to press buttons that CA and EU would have to make jelly buttons illegal.

responding "if you can't be bothered to write it, I can't be bothered to read it" really sets people off

That’s crazy, it seems like such an obviously fair policy to me. They rarely even read their own generated text. I don’t understand this mentality

I don’t even know what Codex is anymore and at this point I’m too afraid to ask. I remember the days when it was a VSCode extension.

Didn’t they also name their new hardware product Codex?

We're 3 to 4 years into the LLM revolution and the rate of progress is still impressive.

The “Attention Is All You Need” paper came out in 2017 and OpenAI released GPT-1 in 2018. ChatGPT was released in late 2022 but that was not the beginning of LLMs. Transformers and generative AI go back even further.

I feel like there’s a takeaway here for the male loneliness epidemic but I’m not sure what

Can’t wait to see all these libertarian venture capitalists begging for another government bailout

This could actually be the fuckup that kills OpenAI as an independent company.

I wonder if they’ll be the Lehman Brothers of this bubble

“Do nothing” can be a decision

aliems

I would guess it’s the same force driving absurd stock valuations — the money supply doubled around COVID and all the new money has to pool up somewhere. Some of it ends up in stocks and real estate, but once those become obviously overvalued it starts pooling up in more fringe investments like trading cards. It’s the same dynamic that created exotic mortgage backed securities that led to the 2008 financial crisis. There’s literally trillions of dollars of capital that’s slowly losing value from inflation and the owners of that capital are desperate to find investments that will preserve or increase their wealth.

Stocks, especially without dividends and negligible voting rights, are basically baseball cards for companies.

There’s some Sanskrit texts that don’t have English translations that I’d really like to read so I was going to use an LLM to create translations. Does anyone know how well LLMs handle Sanskrit or have any suggestions? Are any particular models better than others? Especially because I know that ancient texts sometimes use different dialects of Sanskrit and have different challenges.

The one I’m mostly interested in translating is called the Moksopaya, here’s the Sanskrit on GRETIL: https://gretil.sub.uni-goettingen.de/gretil/1_sanskr/6_sastr...

The general fallacy of the “but inference is profitable” argument is that it tends to ignore all the costs of building and training the model. Given the fact that 1) that’s not trivial, and 2) the arms race underway means one can’t stop training, then it ruins the financial picture.

Or that it’s all hearsay and no one has released financials yet?

Every business wants you to be addicted to their products, whether it’s social media, cigarettes, video games, cosmetics, pharmaceuticals, supplements, fashion, electronics. Even seemingly innocuous things like reading the news or exercising are often addictions, people are running away from their fears and insecurities. It’s a systemic problem in western society, removing infinite scroll is such a ridiculous idea in the face of a core societal problem like this.

My suggestion would be to look at something like the rat park studies. When mice lived in a nice communal space with other mice, they preferred regular ware over drug laced addictive water. I suspect people are similar, they’re not intrinsically prone to addiction, they’re just trying to treat psychological problems like fear, loneliness, insecurity. We need to attack this root of the problem, but there’s so much money in ensuring that people feel empty so they’ll buy shit to fulfill themselves that there’s no actual motivation from anyone with power to do anything. That’s why you get these dumb ideas for laws, they’re not enough to impact the businesses (who are the real constituents) but politicians can still say they’re doing something and something is better than nothing.

No my point is that none of these are by themselves even coercive design, there’s no distinction between a smooth or pleasant UX experience and a coercive design without context.

Should all buttons require a confirmation dialog to prevent people from making decisions too hastily? Should we ensure all interactions have at least 100ms of latency so users don’t get mesmerized by a smooth experience? Maybe we should set a max color saturation so nothing looks too enticing. We also don’t really need box shadows or gradients, they’re clearly meant to mesmerize users into making bad decisions.

What an absurd and pointless precedent to set.

Basically all people with monetary investment in the thing being hyped

and generally not to be trusted

There are many AI bulls who adamantly disagree and cite Tao’s statements about LLMs for mathematical proofs as an example of how advanced and autonomous these systems already are

What kind of repercussions will OpenAI face?

This is only like the 12th reason not to trust OpenAI. The culture starts from the top

Words have more meanings than ever

Right, that's the problem

but last century produced wittgenstein;

I don't think a philosopher who died 80 years ago is driving the change in how words are used in the last 20 years. It has more to do with the Internet and the cultural forces driving people to use hyperbole or make things up to make money in the attention market. This article wouldn't be on HN if it was just about Dua Lipa starting a bookstore, they added "banned" so it would catch peoples attention even if that's basically a lie.

perfectly clear communication was always a polite fiction.

The comment I replied to is trying to argue that it's ok to call books "banned" even if they're not banned, because it's like the term "cancelled" which at one point meant someone whose content was actually cancelled but I guess they're suggesting it doesn't mean that anymore either.

I'm not arguing that words should have perfect meanings, that is obviously a strawman, but this article and comment thread are using words to mean the complete opposite of their common meaning.

I miss the days when words still had meaning

https://xcancel.com/bryan_johnson/status/2072069730517860385

The power of network effects. Same with Elon and Twitter

Exactly, this is just application of property rights and enforcing existing consumer protection rights. You can’t tell someone they’re “buying” something if they don’t own it. Imagine someone doing that with house rentals or car leases, it would obviously be considered fraud.