This is true, but it also assumes continued investment and steady progress.

We had been making some headway in terms of clean energy before the current administration started undoing a lot of that progress. And there is no telling what the next administration will do.

> Pregnant person can do whatever they want with their body.

Yes, they can. But whatever they are doing to their body, they are doing to their baby as well.

Can they drink and smoke while they are pregnant? Sure, it is their choice.

Should they? Perhaps not, if they care about the welfare of the little human that is growing inside of them.

I'm trying to see how this could be net positive.

it's a net positive for robinhood, not the trader necessarily

Net good is kinda uncalculatable here because many of the negatives are a little subjective (I'm thinking AI stock traders x-risk) imo

The article specifically mentions at least one property owner who has been denied any recourse because of the lack of before/after photos (presumably before that specific rental).

How do you rent out your place and don’t take any photos before?

I’m sure Airbnb operators get comfortable turning it over every few days without having to constantly take photos. Most guests don’t bring robots in to smash up the dishwasher and dent the walls

Sure but taking a photo of everything in its “good” state seems reasonable. You don’t need to constantly update the photos if nothing gets smashed.

We don’t really know the details. Perhaps they had “good state” photos that were rejected by Airbnb for not being recent enough.

It's easy to say that as someone who doesn't actually rent out their property regularly on airbnb.

Out of curiosity, what brings you to Hacker News?

I programmed for 15 years and I like watching the dumpster fire. This is one of the better managed forums on the internet, so even though I don’t do that job anymore it’s nice to check in.

(there's an email on my profile if you'd be open to chatting about it sometime. I only ask because I'm curious about making a similar transition)

The way that Mythos is likely being used to train these publicly available models, I wonder if there will always be a private, mostly/wholly internal model that is significantly ahead technically but is reserved for internal or "VIP" use.

If there isn’t they’ve obviously missed an important and lucrative market.

In fact, there should be more and more secret tiers for bigger and bigger money.

There's really no evidence that OpenAI has shared everything they have.

Confidently yes. OpenAI for sure has been training larger models internally and distilling.

Pre-training scaling laws all support larger models being more cost effeceint to train then smaller models. And distillation is comparably cheap. So you can get the most juice by training the biggest model you can and distilling it.

And there seems to be a ton of experts on the opposite side.

As they say, the truth tends to be somewhere in the middle.

Did you test this before shipping?

It errors out when I give it a non-paywalled article.

Yes you can see all the working examples that people have tried just below the input.

I'm not doing anything fancy with the scraping so even if no paywall but some botblocking it will probably fail but worked across bbc, cnn, politco, and several other major news sites I tried

Paywall

While there is a meaningful subset of the population that will do this, I wonder if the vast majority of people, even if they share the same negative sentiment towards AI and in particular how Google is pushing it in their products, will find the friction in changing up their workflows and routines to be too much to actually change things up.

While Google has certainly failed with their products before, by and large as a company they are "too big to fail" at this point so to speak where the myriad products they offer have become a very significant part of the digital infrastructure on which much of the world runs.

Anecdotally I've also switched to DuckDuckGo in the past because I didn't want Google looking over my shoulder on my browsing history (this was pre-AI), but I ended up coming back because I felt the search results weren't quite as good (or perhaps it was just in my head, and the differing UI was enough to throw me off).

And this may be a contrarian opinion, and while I hate the idea of Google mining all of my data and monetizing it, I actually find value in the AI 'previews' that Google provides (and will often ask it follow up questions as a means of getting 'free' LLM responses back for 'easier' prompts that I don't want to burn Claude tokens on).

> weren't quite as good

A lot of people keep saying this, and yet I never saw a reasonable example. Whenever DDG fails for me, the Google's results are even worse.