Hacker Newsnew | past | comments | ask | show | jobs | submit | itake's commentslogin

More accurate and smaller.

quantization = process to make the model smaller (lossy)

dynamic = being smarter about the information loss, so less information is lost


Thanks, that makes sense.

I find I leave the 'important' apps open longer, trying to recheck if I got a notification.



What better alternatives do we have? Not tying my account to a phone number, but rather saving thirteen words, is exactly the UX I've always desired. I don't even need privacy, but I hate losing things when I inevitably lose my phone number.


Ricochet (chat via Tor .onion circuits): https://www.ricochetrefresh.net/

Tox (if you're addicted to phones): https://tox.chat/


>What better alternatives do we have?

Set up your own XMPP or Matrix server and only expose it via Tor.


I think it was xbox?



Perhaps you're thinking of Bunnie Huang, who wrote a book about reversing the xbox. I love Bunnie because he seems to be in it for the joy and the sharing of information.

Geohot (IIUC) hacked the iphone because apple didn't allow devs to run their own code at launch, and the playstation because sony removed the ability to run linux on the console. I love geohot because he seems to be in it to stick it to the man.


I think his iphone hack only involved hacking the cellular chip, so that it would work with networks other than Cingular/AT&T.

Then a Russian guy traded him a BMW for it.


I'm curious how the employees faired. Seems like they may bet getting nothing out of the deal if the investors get their money back.


These days, most employees getting nothing out of the deal is par for the course for acquisitions, unfortunately. The acquisition price is almost never exchanged directly for shares in the company as implied, often a chunk of it is kept for key personnel retention, etc. Typically just enough goes towards the share purchase to make investors happy, and the rest is structured as incentives for founders and key execs with milestone payouts. That‘s the set of people with leverage towards making the acquisition happen, so that‘s who gets paid.

If you‘re just a regular employee with some options, and the acquirer doesn‘t want to keep you on, you should expect nothing.


> Typically just enough goes towards the share purchase to make investors happy, and the rest is structured as incentives for founders and key execs with milestone payouts.

So they're getting the employees' shares without compensating the employees?

And there's incentives paid to the people who approved the deal, separate from their shares?

(I've heard of liquidation preferences, but never by the person making a job offer with stock options. Bribery also never came up.)


Yes, and yes. The sibling comment here about liquidation preferences is correct, and these separate incentives are usually structured as retention incentives — eg, compensation for future work with the acquiring company.

Shareholders are of course free to sue the board for acting outside of the interests of the shareholders overall, but this happens very rarely because typically the company would otherwise be shutting down and it’s very hard to make the argument that the deal undervalues common shareholders’ shares.


Because “shares” are not all the same. Preferred vs common, so unless you negotiated some kind of preferred share terms, assume your shares are worthless. For a non publicly listed company. For a publicly listed company, the details are all publicly available, so the different types of shares will have their different prices be easily available to see.


If that's true, when a startup is making you an offer for ISOs of common shares, and explaining it... how likely are they to know that, in event of a successful exit for the startup, your shares would be diluted and preferenced to 0 value?

(The two most recent offer equity components I accepted were "2%" and "a million shares". On the latter, an upper exec did a kind of deal-closer meeting for their offer, showing me a spreadsheet, estimating how much the options would be worth if there were an exit in X years at $Y valuation.)


> If that's true, when a startup is making you an offer for ISOs of common shares, and explaining it... how likely are they to know that, in event of a successful exit for the startup, your shares would be diluted and preferenced to 0 value?

If they have any experience, or even just browse a forum like this, they should be 100% likely to know. The person on the opposite side of the negotiating table has a goal of giving you as little as possible in exchange for your work (and vice versa).


Brex has been around for a long time, so employees will have been issued stock options with vastly different exercise prices.

Early employees' options will have value, but more recent options are likely underwater.


I strongly suspect they shifted to RSUs at those valuations.


It seems unlikely that regular employees would be issued RSUs. Tax is due at vest, and you can't liquidate to fund the tax bill.


There's double trigger RSUs and so on that allow you to have reasonable tax treatment, due to the theoretical threat of loss if liquidity isn't available. I worked at a company that had this at least while I was there.


I was a regular SDE at brex for a couple years and my various documents about comp say I have RSUs, and carta says so as well.

I've never bothered to understand the details since none of the private companies I've worked for have had the non-cash portion of their comp be worth anything but $0 before.


The usual move here is "double trigger" RSUs that don't vest until a liquidity event, thus no taxes due until said liquidity event.


Right. Plus often the tax is paid out of RSUs given, you just get less in RSUs, some is subtracted to pay tax.


Are those common for regular employees?


highly common


Nope, you as a company owner are highly motivated to shift to RSUs once you hit a certain valuation and number of employees. Everyone does it.


Can you expand on why at a certain valuation and size you would shift?


options can only last 7 years, but brex was founded in 2017.


Who says options can only last 7 years?


ISO options have to expire within 10 years of when they are granted. Sometimes companies make them expire earlier than that, so OP might be thinking of options they were granted. E.g. I once had options that expired 30 days after ending employment even thought the ISO requirement is up to 90 days.


When the options expire do they give new equivalent ones to the employees that hung on? Otherwise what’s the point?


You have to exercise the options or let them expire. You normally have 10 years not 7, but if a company comes up on 10 years after they issued their first options, they might try a tender offer to buy some employee shares. If your 10 year old "start up" shares can't be sold anywhere, then they probably aren't worth exercising. A company that can't provide liquidity to employees for 10 years will probably never do it.


You can exercise the options before that time is up, paying the strike price to convert them to shares


But then you’re potentially stuck with worthless options you can never trade, right? Seems very unfriendly to employees


> Seems very unfriendly to employees

Ding ding ding ding ding!

Most ICs figure this out sooner or later. Unfortunately many only discover it after being screwed hard financially.


While I don't think it's the case here, but a lot of time there is more liquidity preference than the deal value so employees can only get what investor want them to pay.


We know how the employees did. Same as ever. They got whatever slop was left in the trough after the big pigs ate their share.

Bitter about VCs? Me? Never.


You can't eat private insurance.

The consequences of not being able to produce enough calories is severe. It is much better to overproduce and everyone gets fed than producing just enough and a climate event erases out 20% of our calorie production.


The US produces an unbelievably enormous calorie surplus way beyond what is needed for the health of the country and in fact its detrimental.

The biggest is not even used as food, over half of corn acreage is used for ethanol. That's an amount of land that's truly beyond comprehension. Its a horrible program as well, corn ethanol is worse than the gasoline it replaces in terms of carbon footprint when taking land use into account. And it raises the price of food. And we even subsidize it multiple times, we subsidize the crop as corn and then we subsidize it as ethanol. Biodiesel and renewable diesel (different products) have spiked in recent years as well, most of that is made from soy, canola, or corn oil. They have similar problems though aren't as bad as corn ethanol.

Another huge negative surplus is the amount of liquid calories, mainly soda, that are consumed. Most nutrition science that I've read points to the enormous amount of liquid calories as the part of the US diet that is driving obesity epidemic. There are of course other aspects to the obesity as well.

Finally, substituting some of the US consumption of beef with chicken and some of the chicken with beans.

To recap US overproduces calories to the point that it hurts the country. It damages the land, the ocean with dead zones, the climate with carbon. We pay for it multiple times in subsidies and with higher food prices. It hurts our health which we pay for in suffering, shortened lives and health costs.


> over half of corn acreage is used for ethanol.

That doesn't mean much without more details. Corn is used as a tool in the crop rotation to enable growing foods for humans to eat. As we learned before ethanol's time in the sun, farmers are going to grow it anyway to support their rotations. The only question is if it is better to recapture that into usable energy or to let it rot out in the field.

> ... when taking land use into account

But if not taken into account? The harsh reality is that ethanol plants are unable to pay cost-of-production-level prices for corn. It now typically costs $5+ to produce a bushel of corn, while ethanol plants generally start to lose money as the price rises above $4.50 per bushel. You're not growing corn for ethanol. You accept selling corn to ethanol buyers when you can't find a better home for it.

Corn especially is a tough one to predict. A couple of years ago yields around here were nearly 100 bushels per acre higher than normal! Even if we put in the mightiest effort to grow exactly the right amount of corn for reasonable food uses, that 100 bushel surprise means a good 1/3 of your crop has no predetermined home right there. Of course, it can go the other way too. If you end up 100 bushels per acre short of what you expected...

Between needing to grow extra to protect against unexpected low yields, combined with unexpected high yields, half to the corn crop having no home (and therefore ending up as ethanol) isn't that far outside of what cannot be reasonably controlled for.

> To recap US overproduces calories to the point that it hurts the country.

That's fair. We don't have the technology to do better, unfortunately. Maybe once LLMs free up software developers once and for all they can turn their focus towards solving this problem?


> Corn is used as a tool in the crop rotation to enable growing foods for humans to eat. As we learned before ethanol's time in the sun, farmers are going to grow it anyway to support their rotations. The only question is if it is better to recapture that into usable energy or to let it rot out in the field.

This doesn't make sense in light of the large expansion of corn acreage that corresponds to ethanol policy.

https://afdc.energy.gov/data/10323 https://ycharts.com/indicators/us_corn_acres_planted


> large expansion of corn acreage

Corn acres have expanded, but the same is also true of other crops. Given corn's role in the crop rotation, it stands to reason that when other crops expand, corn comes with it. There are a lot more mouths to feed nowadays. The world's population has grown by approximately 30% since the last change in ethanol policy.

> that corresponds to ethanol policy.

The ethanol subsidy in the last policy change ended in 2012, yet, as you point out, corn acres have continued to expand, which seems contrary to what you are trying to suggest. What specific correspondence are you finding?


corn->ethanol is government subsidized robbery. I paid for those nutrients let them rot. Now consumers have to buy more and eat more calories to get the same nutrition. All so we can have net negative ROEI ethanol?


> corn->ethanol is government subsidized robbery.

There was that brief period where subsidies were enacted to spur on construction of ethanol plants to take up the excess corn that was rotting leading up to that time. They have long since come to an end. You could still call E10 requirements a subsidy, but you'd only be paying for that if you willingly chose to consume the product.

> Now consumers have to buy more and eat more calories to get the same nutrition.

Why?

> All so we can have net negative ROEI ethanol?

I'm not sure your math is mathing. Recapturing something, even with some marginal loss, is still a greater return than nothing.


Your buffer here is meat. Cattle are tremendously inefficient consumers of grain. Eat your burgers in the bountiful years, then slaughter 75% of the herd in a hardship year, eat well for six months, then spend the next three, four, five years eating more grains while the herds recover.

Ethanol is another one.

That's the sensible way to do it.

Somehow I doubt that it's the way we do it... But maybe the variability is coming from world trade and developing nations.


Cattle are inefficient consumers of grain, but highly efficient consumers of grass. Most land used for pasture can't effectively be used for anything else.


This argument might sound good, but those cattle are fed crops, not just sunshine and the grass they walk on.

Most crops grown in the US are used as animal feed. They are dependent on arable land that could be used to grow food for humans directly and much more efficiently. We just like the taste, so we accept the inefficiency.


Eh. The "inefficent calorie conversion" take is sort of lazy and misses the nuances. I just looked it uo, and it seems that only about 55% of yields are for feed, and there is definitely some more nuance there, since a lot of feed meal comes from stalks and parts if the plants humans would not consume. This notion of calorie inefficiency also misses the mark on what would be planted and harvested instead to contain the same bioavailable nutrient profile thay comes from meat. In otber words, using land for feed to convert grains to another type of food is probably more necessary than just "taste".

I don't care to research it further, but I own a small 5 acre farm and can attest that some crops grow in some areas and some don't. So even if you did map it all out on a piece of paper where you'd get all your beans and lentils and whatnots I doubt it would work in real life. Cattle can handle a couple hard freezes. My tomatoes can't.


You’re right that a lot of livestock feed is crop residues/byproducts humans don’t eat—but that doesn’t make beef “necessary” or erase the land/opportunity-cost problem. Globally, ~36% of crop calories go to animal feed and only ~12% of those feed calories come back as animal-product calories (Cassidy et al.). Livestock still consume ~1/3 of global cereal production (Mottet et al.). And in full-system LCAs that include grazing + feed land, meat/dairy provide ~18% of calories and ~37% of protein but use ~83% of farmland; cutting them can reduce farmland >75% while still feeding the world (Poore & Nemecek / Oxford). Plus, even if pasture isn’t croppable, it can be restored—land used for animal foods has a big carbon opportunity cost (Hayek et al.). Nutritionally, major dietetic bodies say well-planned vegetarian/vegan diets can be nutritionally adequate, with attention to nutrients like B12.


There is, as you say a lot of nuance here. Making cattle go away doesn't suddenly make say 55% more wheat suddenly appear on market shelves.

Indeed the argument to remove beef production has always struck me as an interesting starting point to a longer conversation.

So ok, cattle are gone, and there's now say 30% more grain on the market. Presumably this lowers prices to humans? Do people suddenly eat 30% more bread?

Health, and weight, issues aside (not sure an increase in carbs at the expense of protein is a win), do people just shift to other protein (like chicken). Does this mean a huge oversupply of grain, and a consequent drop in prices?

Let me put it another way. Does removing a market currently consuming 30-50% of the crop make things better or worse for farmers?

IMO Having livestock feed as a market keeps prices up, and as this article points out they're still too low. Killing off the cattle market kills off grain farmers too. I'm not sure that's the win people think it is.


We feed the average cow >10lbs of grain and also some alfalfa for every pound of meat we get out right now.

Part of the cull would likely be shifting towards more grass fed production. Another part would simply be prioritizing chicken or pork for a while.


Americans would riot without burgers.


Why AUD and not SGD?

I guess AUD has higher GDP (but much lower GDP per capita and worse forex rate to the us dollar)


> Why AUD and not SGD?

Fair question. And in some circles people would substitute AUD for CHF, and SGD is essentially in the same position as CHF (small, stable, well-connected country).

But AUD mostly because Australia is commodity-rich and so AUD has a naturally high exposure to international trade especially in Asia-Pacific region. Its a bit like CAD but the problem with CAD is its neighbour. ;)


IMHO, instead of 90-day memory loss, use an LRU eviction strategy.


My thought was the image metadata already includes the datetime... but I guess you're right that filesystems wouldn't support sorting by that.


FAISS runs in RAM. If your dataset can't fit into ram, FAISS is not the right tool.


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

Search: