Hacker Newsnew | past | comments | ask | show | jobs | submit | a-r-t's commentslogin

Nice to see a Festool miter saw in his shop, Lynch knew what he was doing.


There was an auction of a lot of his memorabilia a few months ago, it included a lot of Festool stuff. He was an avid woodworker (the sale also included furniture he made). I like how the work table where you can see the miter saw is made of the most utilitarian plywood, it feels like he was working until his last days


There is a good Veritasium episode on her last flight going deep into technical details of what went wrong: https://www.youtube.com/watch?v=zTDFhWWPZ4Q


Yes, the Veritasium episode is great.

In short: there's plenty of evidence Amelia Earhart was reckless. I'm sad that she paid with her life, but that is sometimes what happens when you're reckless while using dangerous machines.


Captain A. G. Lamplugh, a British pilot from the early days of aviation once famously said “Aviation in itself is not inherently dangerous. But to an even greater degree than the sea, it is terribly unforgiving of any carelessness, incapacity or neglect.”


> Aviation in itself is not inherently dangerous.

Yes, it is. Otherwise, "any carelessness, incapacity or neglect" wouldn't be so "terribly unforgiving".


Agreed. It is like saying a tightrope isn't trying to harm you, just don't fall off.


In car transit (like most things in life), you can do everything right and still die. In aviation, if you do everything right, you'll land safely.


> In car transit (like most things in life), you can do everything right and still die.

Same with aviation. The DHL Flight 611 over Überlingen, Iran Air Flight 655, Ukraine International Airlines Flight 752, American Airlines Flight 5342, Pan Am Flight 1736 are just the few easy ones which comes to mind immediately.

> if you do everything right, you'll land safely.

You. And the people who designed your aircraft. The people who maintain your aircraft. And the ATC. And other pilots. And the people on the ground operating anti-aircraft missiles.


The two pilots in the Tenerife aircrash would beg to differ.


>> In aviation, if you do everything right, you'll land safely.

The pilots who died in the 787 MAX crashes would disagree. They did everything exactly as they were trained to do and still crashed.


> 787 MAX

You are thinking of the 737 MAX [0]. There is no such thing as a 787 MAX yet [1].

[0] https://en.m.wikipedia.org/wiki/Boeing_737_MAX

[1] https://en.m.wikipedia.org/wiki/Boeing_787_Dreamliner


Yes obviously, sorry that was a typo.


Yep, and how my instructor said, what's worse is that the sky can let you enjoy being careless 100 times and then punish.


Same thing with farmers: it's usually the old, experienced farmers who die in dumb ways. They've been doing the same dangerous thing their whole lives and become complacent until it catches up with them.


Interesting, didn't know about farmers. Skydivers are the same, most accidents happen to experienced ones. But it's understandable, as adrenaline wears out with experience.

Pilots typically are trained against that complacency, plus, as they say, everyone can be stupid for 15 minutes a day, plan for that. I found piloting pretty boring, if done right. Talk about soul-crushing.


That's just life.


Yeah this was a great video, so many errors.

The experienced navigator refusing to fly with her was correct, but I do wonder if he had been there if he would have been smart enough to save them.


Probably . . . from what I have read in the past, a better understanding of radio direction finding probably would have been enough to get them to Howland Island.


If we are speculating here, why not just go straight to an LLM serving all requests directly? No code needed.



There's new information posted on the internet everyday though. Press releases, earnings results, new legislation or supreme court decisions, sports scores/results, new technologies and discoveries. The original "thing that happens" isn't produced by bots.


Hi Jeff, are there any plans to support dual-channel audio recordings (e.g., Twilio phone call audio) for speech-to-text models? Currently, we have to either process each channel separately and lose conversational context, or merge channels and lose speaker identification.


this has been coming up often recently. nothing to announce yet, but when enough developers ask for it, we'll build it into the model's training

diarization is also a feature we plan to add


Glad to hear it's on your radar. I'd imagine phone call transcription is a significant use case.


I’m not entirely sure what you mean but twilio recordings supports dual channels already


Transcribing Twilio's dual-channel recordings using OpenAI's speech-to-text while preserving channel identification.


Oh I see what you mean that would be a neat feature. Assuming you can get timestamps though it should be trivial to work around the issue?


There are two options that I know of:

1. Merge both channels into one (this is what Whisper does with dual-channel recordings), then map transcription timestamps back to the original channels. This works only when speakers don't talk over each other, which is often not the case.

2. Transcribe each channel separately, then merge the transcripts. This preserves perfect channel identification but removes valuable conversational context (e.g., Speaker A asks a question, Speaker B answers incomprehensively) that helps model's accuracy.

So yes, there are two technically trivial solutions, but you either get somewhat inaccurate channel identification or degraded transcription quality. A better solution would be a model trained to accept an additional token indicating the channel ID, preserving it in the output while benefiting from the context of both channels.


(2) is also significantly harder with these new models as they don’t support word timestamps like WHISPR.

see > Other parameters, such as timestamp_granularities, require verbose_json output and are therefore only available when using whisper-1.


The MS announcement is limited to scams/phishing. Google mentions both scams and spam, but somehow I still get 15-20 spam emails a day that even the smallest LLM should be able to classify correctly.


You mentioned Django, but I couldn't find much on your website about how it works with your forms. Will I have to manually replicate all form fields on the client side? Will it work with Django form validation and show errors? I'd suggest creating a documentation page for each framework you mentioned to explain how it will work together.


The goal from the start was to raise $230M and then get acqui-hired for $116M by HP?


https://support.humane.com/hc/en-us/articles/34374173951373-...

"Your Ai Pin will continue to function normally until 12pm PST on February 28, 2025. After this date, it will no longer connect to Humane’s servers, and .Center access will be fully retired."


It's a $700 device with a $24/month subscription fee, and will be bricked in ten days.

At least the equally-panned Rabbit R1, released around the same timeframe and riding the same AI hype, only cost $200 with no subscription.


Amusingly, Rabbit just announced today that they're now shipping at Best Buy - their first steps towards brick and mortar exposure.

https://www.bestbuy.com/site/rabbit-r1-mobile-ai-device-pers...


> Our support team is here to help at support@humane.com through February 28th, 2025.

If someone had written this as satire when they released the pin it would have seemed a little harsh


Alas, the pricey e-waste reaches its unsurprising end of life.


Off topic, but the font in the article is hard on the eyes.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: