> Waymo says it has 170 million miles and I found 2 fatal accidents it was involved in (fault doesn't matter as fault is not included in the other stat either)
fault absolutely matters. and with two fatal accidents, it should be trivial to find out if waymo is at fault.
You could have looked up the collisions instead of making vague accusations of bad faith. In one, a Tesla going 98MPH collided with a line of vehicles (including a Waymo) stopped at a light. In the other, a motorcyclist hit a Waymo from behind and fell off their bike, where they were struck and killed by another vehicle (that fled).
I never went (just don't like it), but for a long time it was a primary source for a lot of important things so one kinda gets exposed to it anyway indirectly and / or through embeds. But that's the filtered version, anyway.
I made the experiment some times ago: it's faster to edit the link, replace x.com with nitter.net and load the content from there, than wait until the original page load. The time difference isn't meaningful (it's less than a second), but the fact that a website loses against a manual redirection to another front end is still a spectacular proof of incompetence.
I wonder if it is by design. On Nitter, which is fast-loading and streamlined, it is trivial to ctrl-click on a bunch of commenter accounts and readily see they are bots (photo from This Person Does Not Exist, unrealistically single-issue posting, etc.). On X where things move slow as treacle, it is harder for the average person to figure that out.
Same here. There is always the "log in" nagging pop-up. But I will never log in again to the right-arm-gesture guy, so this is just time wasted for me.
We could need some decentralized way to share "tweets" though. Not linked to greedy billionaires.
Being able to use AI to generate the steps to synthesize proteins means that you can use it to use it to generate the steps to synthesize known toxins. Suddenly, once difficult to attain knowledge is now available to everyone.
Still need to do it after getting those instructions. Not to forget equipment and precursors. I think just getting list of steps won't make it too much easier. Getting it mostly right is quite hard in many cases. And then with trivial cases you wouldn't even need AI. But just find something already documented.
Concur entirely. After reading _The Making of the Atomic Bomb_ (highly recommended, BTW) you know the steps to make a U-235 enriched atomic bomb. The difficulty comes from obtaining the enriched uranium.
I've heard it as shorthand, in-context. Never heard it being chat-as-in-kleenex. I've supposedly been in the industry since before ChatGPT. Maybe that's where I went wrong :D
in the us, entirely legal. you may get law enforcement knocking on your door with questions, from time to time.
best to do it as part of a non profit or other organization, and being open about running a node, so that law enforcement/etc knows what they're dealing with
> you may get law enforcement knocking on your door with questions
If you host an exit node on your home internet connection, the daily thing you will struggle with is not law enforcement, but rather the fact that your public IP will be blocked by most big websites to the point that simply browsing the web will become a pain.
Then whoever did this experiment will likely shut down the node and reset their router to pull a new IP, which could be reassigned in short order, thereby causing trouble for an innocent neighbor.
> best to do it as part of a non profit or other organization
I've always thought that universities would be a great safe haven for exit nodes, probably managed by a student club but using campus's infra (with their fully informed permission and support, of course).
I think this would make them less likely to be shut down by the police because a university would lend them more credibility as a civic-driven research project rather than some random person running "shady" stuff, as most people unfortunately assume.
Here in Mexico there's very few Tor nodes running at all, but some of the longest running (non exit, TIL) nodes are run by UNAM [1], which is the biggest university.
That being said, I studied in a different university and I didn't have the confidence or knowledge to propose anything similar, so I've no idea of how effective would such a project be or how hard would it be to run.
They are. In general, having a big and respected institution to hide behind does wonders for avoiding suspicion. It would have to be formally approved by whoever is responsible for these things at the university - you couldn't just plug it into a random port (well you could but you might get Aaron Swartzed).
Personally, I just think that LLMs are an artificial line to draw based on current personal preference.
I'm not an AI maximalist, but I really can't see the issue with using it to preserve this hardware (especially into RTL for FPGA), if it's highly accurate.
These machines are dying in real time, and having cycle and operationally accurate emulation options is more important than the bad taste the Glorified Copymachine leaves in my mouth, to be honest.
Honestly, by quantity alone, I've seen more impressive and quality reverse-engineering and documentation work of these platforms done in the last year than i have in the last 20.
Hardly, the country is still heavily cash-using, and where transitioning to electronic payments the local PayPay system seems to be popular. Suica payments for stores are basically only for dense areas of Tokyo.
Hardly? Credit cards passed cash for retail purchases last year. They're neck and neck.
If you break down by age demographic, you'll probably guess who's using what. And the ones using cash are quickly disappearing.
And let's say it wasn't, and if cash was still king: 10% of Japan's payments infra being weaponized during tariff negotiations would still be a huge dent to their economy. The country had been migrating over heavily, but Trump 2 actions put the brakes on a lot of EMV rollout.
PayPay is an evolution/merge of Line Pay (long story), which struggled to capture market share in Japan past peer-to-peer payments, due to a bunch of factors. You'll see it in stores, of course, but it's adoption isn't amazing (not bad, just not amazing). Suica benefits from FeliCa/NFC-F lock-in, and considerable existing deployment, plus integration with the rail network. Almost everyone in Japan has a FeliCa/NFC-F device with currency loaded.
FWIW I am an American, but I am working on payments infrastructure and billing for a American owned subsidiary of a Japanese corporation. This happens to be a subject that I am wading into on a daily basis, at least from a technical side.
Oh, and you might be interested to hear what's happening in the UK with OysterCard. They recently finished their EMV rollout and then paused all OysterCard sunset timelines. I wonder why.
it's widely used but in "industry" applications. so ffmpeg is probably being used in a lot of offices (studios) and maybe even being included in end user software.
The file format that we're describing was part of a proprietary embedded SDK for the PS2. The majority of professional integrations would be using the Sony SDK.
Today, a lot of tooling uses FFMPEG for accessibility of those formats. But I wouldn't say it's the majority.
Hence why I said "(studios)" after offices, and "probably" and not "majorly":
> probably being used in a lot of offices (studios)
last place i worked at had internal telemetry on end user machines. i dont know the way it was being used, although i suspect it was a document viewer.
fault absolutely matters. and with two fatal accidents, it should be trivial to find out if waymo is at fault.
it is bad faith, otherwise.
reply