nightsd01

As an iOS engineer I am actually pretty excited about how this will improve manual testing on simulator.

Honestly this is a good thing, and we should ban online sports and casino betting in the US too. They add absolutely zero value to society, they just cause harm.

Studies have shown that most of the profit from online sports gambling for example comes from people with gambling addictions. These companies are just greedily sucking away money from vulnerable people. These companies really do ruin lives and destroy families.

I can see that maybe prediction markets might offer at least SOME value to society, so if I were made global dictator tomorrow, I wouldn’t ban them right away, but I’d definitely want to see more studies doing a risk benefit analysis.

I have to say, I am quite frustrated with Anthropic lately. I so badly want to use Fable to work on a side project of mine, which I used to do previously with no issues. But lately, they must have made some classifier change because it keeps hitting their stupid, overly-hyper-aggressive safeguard due to 'general_harms'.

Guys, listen to your feedback please. I hadn't used OpenAI products in quite a while until this issue came around. They seem to have MUCH smarter safeguards than Anthropic does.

I will never understand why NASA doesn't build at least two copies of expensive space telescopes like this. One single launch failure and boom (literally), the entire project is a failure?

It wouldn't double the cost assuming they do it with even a modicum of efficiency and intelligence. And hey - assuming that the launch goes smoothly, you can get double the benefit

> Since then I've done Opus and then Fable bug / security sweeps regularly.

This part surprises me a bit. Fable always falls back to opus for me the instant 'security' is involved (thanks for that, Anthropic)

I was a bit triggered by your China comment, if only given the fact that there is a very specific institution that drove many formerly prosperous cities and communities into abject poverty (the CCP in Mao's era) and then tries to claim all the credit for the market driving how many people they've lifted out of poverty.

And nowadays with Hong Kong (and probably soon Taiwan) they are proving they are perfectly happy to destroy economic growth as long as it benefits The Party

Plenty of people are making money from AI? And I am not sure I've seen it "making things worse for most people" yet, do you have a source behind this?

If your source is AI layoffs, there were plenty of layoffs with the invention of the horseless carriage, but that doesn't mean it made humans worse off overall.

There are reasons to be skeptical about progress and AI and all that but the 'making things worse for most people' thing you mentioned seems yet to be based in any reality

Got mine on launch day. It's been sitting in the case for over a year. It's frankly just not comfortable to use for more than 10 minutes.

The sad truth is - getting on a private jet would be just about the single worst thing you could do during a nuclear war. Most modern private jets are largely based on flight control systems based on computers and electronics that would be quickly fried by the EMP flash from nuclear weapons even a large distance away. If you wanna die - get on a non-military aircraft during a nuclear exchange

It's doubled because it practically couldn't get any worse. Let's see what happens (sadly) now that USAID has been dismantled

I recently built a drone from the ground up - learned how to build PCB's with the ESP32, wrote all of the flight firmware, etc. and built a controller iOS app.

Extended Kalman Filters are even more interesting because they let you do sensor fusion and such

Why is it exactly, that mobile providers can’t catch this…? Just like with SMS fraud. It might cost them a few cents per subscriber to do effective anti-spam measures, but now society has to pay the cost

I am not an expert in AI by any means but I think I know enough about it to comment on one thing: there was an interesting paper not too long ago that showed if you train a randomly-initialized model from scratch on questions, like a bank of physics questions & answers, models will end up with much higher quality if you teach it the simple physics questions first, and then move up to more complex physics questions. This shows that in some ways, these large language models really do learn like we do.

I think the next steps will be more along this vain of thinking. Treating all training data the same is a mistake. Some data is significantly more valuable to developing an intelligent model than most other training data, even when you pass quality filters. I think we need to revisit how we 'train' these models in the first place, and come up with a more intelligent/interactive system of doing so

Apple’s biggest problem is their commitment to privacy. Delivering effective AI requires a substantial amount of user data that Apple doesn’t collect.

Their other problem is they value designers and product managers more than engineers (especially top tier AI engineers).

Both problems are basically the death knell of any hope for Apple to have good AI, but combined? It’s never gonna happen. Which is sad because Apple’s on-device hardware is quite good.