voisin

For the uninitiated, what technologies do Waymos use that allow them to offer unsupervised FSD?

I ask because I find my Tesla Model Y FSD to be incredible, requiring effectively no intervention (only in parking lots because I am picky about where I like to park), and I often wonder if it is purely bureaucratic inertia that is preventing it from being treated as unsupervised.

The real heart of what I am getting at: will Waymo in Canada nudge the regulators to reconsider Tesla FSD or does it have something special that Tesla is lacking?

Yes

Pothole avoidance is on their “coming soon” list in the last few software update notes. It can’t come soon enough!

I concur. Only once you get used to FSD do you realize how the thousands of micro decisions and constant attention demands of driving actually do add friction. Arriving at a destination using FSD completely fresh mentally is really quite something. The longer the drive, the greater the difference.

> The long-tail of edge conditions to handle is really long, and Rivian is only at the beginning.

I still find it insane that the entire established automotive industry saw what Tesla was doing with cameras and incrementally working toward FSD, and still to this day has not even started to collect data (or even install enough instruments to collect the data!) to get onto the same trajectory. Meanwhile the Model Y is the best selling car in the world multiple years in a row.

They’re all hostile.

I think this is due to lazy prompting. It isn’t hard to get an LLM to write concisely, with a logical flow and to be direct with the point you want made. I’d rather read something an LLM has written in this manner than a lot of things I come across written by humans.

Regarding padding out word counts, I see this more often in newspapers and magazines than I do in AI-land. It’s like Netflix shows trying to meet an 8 or 10 episode minimum - horribly boring with unnecessary filler.

Not necessarily a stunt if they’ve taken a meaningful stake. Smaller companies can acquire larger ones. Market cap reflects equity, not enterprise value. If the target has low debt, the deal can be financed with borrowing. The 1980s LBO wave, led by Michael Milken, is a clear precedent.

It is pretty obvious to me that Anthropic wasn’t prepared with sufficient infrastructure to handle the wave of OpenAI/DoD refugees. Now everyone is getting throttled excessively and Claude is essentially unusable beyond chatting. Their big new release of Cowork is even worse than Claude Code for blasting through session limits.

I am tired of all the astroturf articles meant to blame the user with “tips” for using fewer tokens. I never had to (still don’t) think of this with Codex, and there has been a massive, obvious decline between Claude 1 month ago and Claude today.

> Annual budget $31.3 million (FY 2024)

If it hasn’t been used in 50 years, is there some other use for the registry or the organization or why hasn’t this been cut yet?

Identity politics was the primary agenda well before occupy wall street. Think about the religious right in the early 90s pushing abortion and gay marriage as primary political issues to drive voters.

> In this case we are the horses.

This is assuming the conclusion. The entire question is whether we are the horses or every other example of humans in the past who found other employment that was inconceivable previous to the technological revolution that rendered their old job irrelevant.

Train Dreams

It’s important to take Vitamin D, as a fat soluble vitamin, with dietary fat during a meal. Something about bile production and absorption.

Also important to take it with Vitamin K.

Source? There have been many articles on HN showing the RDA to be ~10x too low (something like 5,000 IU) and that the daily safety limit to be significantly higher than that (something like 30,000 IU).

Edit: for clarity I am not saying it is impossible to overdose on oral tablets, but rather that with most tablets 400 IU to 1000 IU and the safe limit so much higher than these, it seems like it would be extremely unlikely for someone to be taking 30+ tablets daily. Not impossible, but not easy either.

His rationale seems to be validated by Nvidia following the same strategy, no?

Can you elaborate for those less familiar with the successes vs failures?

Is it possible to import Goodreads history?

I love Fastmail otherwise though. I don’t want to switch away, and I don’t want them to force it down anyone’s throat. I just want an option to integrate or a feature I could turn on (even a paid premium tier). My response to another person explains what the problem is - I think it is a pretty common issue.

I get 3-5 email requests a day for meetings. Those meetings are at various locations (so options like Calendly which are more focused on people doing meetings from a single location don’t work). I need to reply with times that I am available. Often they get back to me and say they aren’t available at those times so how about these times. We iterate until we find a time that works.

All of this is manual right now. I’ve spoken to a lot of colleagues in my industry who have the same pain points. A lot of time is wasted on this.

Something intelligent could take into account where I am going to be right before the time I offer and make sure there is enough time for transit in between. It could warn me if I have a few meetings back to back and might need a break.

I love Fastmail and don’t regret ditching the Gmail backend at all, but I do wish I could have something intelligent like this integrated.

I switched to Fastmail and I desperately want to integrate ChatGPT to intelligently schedule meetings. I spend a tremendous amount of time going back and forth with people manually inputting when I am free. I’d pay more for this integration!

I also use ChatGPT (copy and paste) to rewrite long emails for clarity. I’d love if it had pre-written drafts that I could approve or edit and send…

Progressive taxes just mean that at the highest levels more income is going to redistribution. At all other levels there is still incentive to work.

I think it is far to keep working if you love what you are doing. To filter, there should be an absolute cap on wealth at a few hundred million dollars. This would eliminate the incentive to manipulate politics in favour of yourself, but if you want to keep working you should be doing it for society via charity or taxes on anything additional that is earned.

Have a nice ceremony and present a medal for winning capitalism.

It’s almost like we are not optimizing society for human flourishing.

Modern life is spent indoors whereas historically during our evolution it was outdoors

> it is easy to find 5000iu and 10000iu capsules at a drug store

Not in Canada!

If you hold so rigid to the current rules of grammar, you’ll miss out on some truly beautiful works. Many poems, and almost all of Cormac McCarthy.

Subdivisions of starter homes (defined both by size and price) tend to attract young families

What is the cutout for? Just aesthetics?

> 1. Make a video documenting each piece and its story while she’s still alive. Get her to tell the family history, where items came from, what they meant to her. This preserves what actually matters.

The wild thing is that “what actually matters” likely becomes “what doesn’t matter” after one more generation when people who never met the person in the video inherit the video.

We are all just here for a brief time and yet we (myself certainly included!) cling so hard to attempting to leave a mark.

Most people we know only think about us for a month or so after we’re gone. Only our closest family and friends think about us longer and even then maybe not so many years later.