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

Ultimately I've never used Grok and don't want to support the company / CEO behind it, but even without that the 256k context window as given in this link means I won't look at it.

Model intelligence is great, but more often than not I found my issue is not that I would like 5/10% more intelligence, I'm already not even using the highest mode of thinking on most of my queries, and very important queries are better served by asking different model and ask another one to compare and help me decide.

My "big" issues are mostly centered about context loss and how they explode in flight when that happens sometimes without you or them noticing.

I don't know, depends on everyone's usage I guess, and 256k is not THAT shabby, but that's how I feel about it.


Meta comment to your meta comment : this is a very personal opinion but I cannot stop myself from thinking less about your message (even though I agree with it), just because of that edgy wannabe way to go about naming Microsoft. You want to say Microsoft say that, you want to say Github say that but if you say a vaguely insulting name only to have to specify what you actually mean... It doesn't bring anything to the conversation, except lower the level, and if a company has flaws as big as Microsoft you don't need to resort to childish name calling to expose them. This is my personal opinion.

I thought it brought something to the conversation

> just because of that edgy wannabe way to go about naming Microsoft.

I think it stopped being edgy and transitioned to a symbol of user frustration the moment their CEO tried to address the slop without acknowledging their customer pain.

It is not much different from the micro$oft used here often.


won’t somebody please think of the billion dollar company?

This has nothing to do with the billion dollar company, and I have no lost love for Microsoft.

It's about quality of discourse among us, and Microsoft has nothing to do with that.


what do you think about Micro$haft WinBlowz

It means GH is now full of slop and pointing out it's owned by Microsoft, for those who may not have known or LMS who did not until recently

Merging the failling companies into the other ones is the usual Elon thing, Solar City didn't get acqui-merged into Tesla for its great result.

It's not a "scam" in the traditionnal sense, it's riding the bubble while it's there, stock value is "supposed" to be about the company performance and potential but technically it doesn't have to be, it's about what some people are willing to pay for it (the stock, not the product the company sells) and that's all. That's also why tesla has such a valuation.

You can see it in the comments even here and other thread about this IPO, some people read the numbers, and some have just religious sounding comments about it being the biggest revolution ever or making the history book etc ...

And that's also why they need to keep elon as CEO because in the scenario where they remove it and get the best car company CEO and become a great regular car company that works and ships lots of great car ... Their valuation would be reduced a factor of ten


It’s a scam and the party will be over when Tesla finally bites the dust, and it will. The worldwide trajectory is not in their favor.


What makes you say that? Chinese EVs?


Not parent, but Tesla is currently worth more in the stock market than the ten biggest car seller combined

That's with falling sales, with no new product in sight and their latest release a big dud (despite SpaceX ordering massive amount of it to prop it up), and with a pricing and thus margin that is going to be massively lowered or beaten by Chinese ev. Protectionism by the US can only go so far, the US alone doesn't have a trillion dollar worthy car market.

They're also losing the race in self driving, and the biggest innovation in EV (semi instant charging) is coming from China not from them.

They had a lead and a massively advantageous legal situation and they squandered it, now you add the association of Elon to the brand making it toxic in many places and it's not looking rosy, the brand can be fine but it's valuation at one point or another will have a wake up day.


Tesla's EV is worth about the same as the the EV of the term biggest car companies combined. It is highly speculative, but their current new product is the Semi, which started volume production last month, and they have the robotaxi and Optimus in the pipeline.

One advantage they have over other manufacturers is their willingness to move into (and out of) adjacent verticals to optimize product cost/design/etc.

They could be beaten by EVs from China/etc, but only time will tell.


European truck manufacturers like MAN and Scania have already had electric semi-trucks on the market for years. And they are selling, the Danish transport giant DSV alone has 400 driving around Europe.


That's about what one company ordered last month (370).

https://www.teslarati.com/tesla-massive-order-semi-370-units...

It looks like MAN/Scania combined production last year, including buses, was about 1000.

https://www.sustainabletruckvan.com/traton-scania-man-electr...

Estimates for semi production this year are 5000-15000.

https://sherwood.news/tech/what-we-know-about-teslas-semi-tr...


They're second in the race to self driving after Waymo, they're making the Cybercab, seems to be working out great as their cars will cost way less than Waymo's


No, it's a fine, but the fine doesn't absolve you from fixing it too so it stops. You have this delay to submit a plan for how and on what timeline you will fix it. If you don't do it, or take too long, we will keep fining you, increasingly.

An exemple what how in the old microsoft case they ended up puttin a daily fine for non compliance until microsoft balked back and fixed it (after they tried to act tough and pretended to ignore them).

The end goal ultimately is to get it fixed.


How do they enforce a fine on a Chinese company? What if temu says "up yours"?


I visited Temu from Sweden and clicked on the terms of use, this is the first line:

1.1 These Terms are between you and Whaleco Technology Limited, an Irish company.


you won't be able to sell in the EU market anymore


Doesn’t Temu direct ship to the customer? What if they ship in plain unmarked packaging and keep changing the address of the sender? Is the EU customs peeps just going to start inspecting every single package from China looking for items from Temu? That sounds like a logistical nightmare. This sounds like old school thinking where you can stop whole containers full of stuff from a single supplier.


At some point it's a diplomatic incident and will affect EU-Chinese relationships. Even the Chinese government doesn't want to fuck it up for all Chinese companies just because one of them feels like the rules don't apply to them. It's not like the only goods flowing from China to the EU are cheap trash.


Smuggling isn't a great business model for legitimate companies.


The money has to move from the EU to Temu/Pinduoduo coffers at some point.


What logistic company will ship plain unmarked packages? They simply wouldn't be delivered at all.

> Is the EU customs peeps just going to start inspecting every single package from China looking for items from Temu?

They might, why not. It would be unwise to pick a fight like this for any company.


Temu has EU warehouses they appear to ship from: all return addresses I've seen are EU addresses.


Say they carry on.... How does EU actually stop people ordering from their website and getting items posted to their house?


Maybe going for the money. Forbit EU banks from transferring funds to known Temu accounts.


They'll put them on naughty list that will be enforced by financial institutions, i.e. it will be an infraction for credit card operators to process such a payment. Financial operators have well oiled compliance facilities and the payment won't clear. If Temu won't get the money, they won't ship the parcel. And if they won't ship, then there will be a bit less carcinogens in EU. Good stuff.


Ordering ISPs to DNS block temu would probably be easier and effective enough.

Or maybe getting google and apple to make the app not available in the EU.


There are still borders and customs inspections, that's how.


Just in a poxy country like the UK it's millions of parcels a day delivered across multiple ports mostly inside containers.... It's simply not feasible to check it all, it would cost a ton of money to have enough checkers and not slow down deliveries.


That's a failure of the state to control its borders.

> it would cost a ton of money

That's why the EU is imposing a €3 fixed customs duty per item (and later another €2 handling fee) effective July 1 (should have been much sooner in my opinion) for small packages (under €150), in addition to the VAT.


> That'd be really easy to spot and also fix, most likely

Ah, reminds me of good old "There are only 2 hard problems in computer science: cache invalidation, naming things, and off-by-1 errors."


> Ah, reminds me of good old "There are only 2 hard problems in computer science: cache invalidation, naming things, and off-by-1 errors."

You quip, but LLM KV caching (from the harness side) is quite easy: You get a cache hit on stable prompt prefixes, period. That means you want to keep the prefix stable, and only append at the end of the conversation. Made up example: Don't put the git branch name into the system prompt part (that comes first), as whenever the branch name changes, that'd trigger a cache invalidation of the entire prompt.

Getting this right requires some care to not by accident modify the prefix, basically, and some design on communicating the things that can change (user configuration, working dir, git information, ...).


That sounds like the experience of writing Containerfiles; since steps are cached you want to pull the thing you are iterating on as far down as possible.


All of this work has been done before in different contexts. Memory management with bigger blocks and weaker definitions that change whenever some grad student gets a bright idea.


100%. Since you mention memory management: Generational GC is pretty much the same idea: Keep the stuff that's least likely to change an important property (liveness) together.

Conceptually the underlying general idea is to sort things based on stability if you can avoid recomputing properties of the stable part.


It's even closer to prefix matching on super long strings by chunk


In France, basically every bank say (show in their app and everything) "if we call you and ask anything like code, confirmation, to do an action, anything, end the call and call us back, don't do anything on a call you didn't initiate".

Same in their app eg you try to do a sepa wire to a new recipient and you get a warning "are you on the phone with someone ? did someone ask you to do that ? please call your bank by pressing this button. By the way we will never call you to ask an auth code or to do a wire"


A few UK banks detect that you're on a phone call and show a message like "we've never called you" or "we are not calling you right now" in their app, I think that's really smart.


The amount of behind the scenes work to get that set up seems impressive.


Here is a fun one, my mobile phone company has an account lock along with a pin and OTP over SMS system. In order for me to activate a new device (like an phone upgrade) with eSIM over the phone, I need to unlock my account with account lock, give them the pin over the phone, and read the SMS OTP to the mobile phone rep online. I get doing the account unlock and verbal pin, but I don't get why they ask for the OTP especially when they train us to never share the OTP over the phone. I even asked the rep about it, but he mentioned that you should never share the OTP if you did not initiate the service request. From a security posture point of view I think that stinks. I am not exactly sure how they expect SMS OTP to work in the case where my phone is not functional.


And then we have the national post office sending its notifications from the scammiest-looking domain they could find: noreply@notif-colissimo-laposte.info


In Turkey, if my bank calls me, they also send a push notification telling "We are calling you. The representative's name is $NAME. You can talk safely".


Unfortunately in the US, maybe elsewhere, pharmacies and medical offices have trained the elderly it’s okay to verify their dob when they call. Costco does that when they call and it drives me nuts.


US insurers expect you to click on sms links and log in with your username, password, and 2fa all so you can receive a fucking marketing message.


Why would anyone stick with an insurer that clearly doesn’t give a darn about them?

Just for a discount?


Well, I could go with a different insurer that blatantly and brazenly lies about their network coverage[1], or I could go with one that doesn't, but still doesn't have my doctors in it, or I could go with the other one that people claim consistently denies care and coverage.

Also, if 'Just for a discount' isn't a reason to use them, do you have $3,000 lying around to wire me? If you do, I'll happily switch to a much more expensive insurer that meets my other criteria, and might or might not send me marketing materials disguised as fishing SMS. (I'll let you know if they do.)

---

Insurers aren't banks or ISPs or gas stations. They don't provide a fungible service that is nearly identical from one to the other. You can't 'just switch'. They are both heavily obfuscated, and heavily differentialized, because the healthcare 'market' is obfuscated and heavily balkanized.

And all of them are utter shit, but in different ways, and if you are lucky, you won't discover the ways in which yours is shit.

---

[1] How this isn't a statutory capital crime for anyone with the rank of director and higher, I have no idea. But the fact that the people orchestrating this are permitted in civil society does lead me to believe that maybe we don't live in a just world.


From what it sounds like… you are the one already paying either way.

It’s just one firm takes, instead of more dollars, some other value out of you?

I’m not sure why the latter is more preferable. But if your sticking with it after seriously thinking about it, then that is your choice.


Don’t have much of a choice. Gotta love our system.


I don't disagree with you in general terms, but what % of total commits or code change came from you and the three others leaving? A long lived long tail is great, but it still often dies out when you cut the head


If you’re not that familiar with the project then what’s it to you?

Personally, I think this is great and I’m going to send this to my boss who might have to make a similar decision someday. It seems to me like they could have just sold the product to them but went to lengths to keep it independent; that’s the type of thing we should encourage.

No offense intended but sometimes this type of comment sounds like the other side of the AI psychosis coin, like an anti AI psychosis.


To be fair regarding taxes it should be that you get all exemptions by default and the other side telling/justifying you why you don't qualify, instead of you not getting the things you should because you're not sure how to interpret the law or don't know this or that rule exists. Taxes shouldn't be that difficult, and the US version of it seems to be behind even the one I have here in europe (where my taxes are done "for me" for anything non business owner related). If the government is calculating your taxes anyway, they should just give the number to you instead of asking you a number and you better have the same as us or you're guilty of something.

Feels like a system that is deliberatly made to be more punishing for those who can't afford the help or the education to figure things out.


I would agree with you on most situations (like 1040 personal income taxes especially).

But in this particular tax credit, there's no way for the gov to know automatically what percentage of payroll was spent in qualified R&D expenses, since it's day to day business operations. Which is why we are _forced_ to hire an outside firm and pay them thousands of dollars (when Claude did an even better job), just to analyze how much of our time qualified as R&D expenses.

The problem I have is that I am forced to have to find a firm to do this, and most firms won't even work with companies as small as ours. So then we're stuck and losing out on years of R&D tax credits at the moment, when I really don't need them anymore, to be honest.


Or punishing to those that don’t pay for software and services to the companies that lobby for it to be this way.


i saw a meme once like:

IRS> Pay your taxes!

me> ok how much?

IRS> idk you have to figure it out

me> ...ok

IRS> if you get it wrong you goto jail


I remember a different ending

me> so you don’t know how much I owe?

IRS> no, we do…

me> ...ok

IRS> if you get it wrong you goto jail


Not parent, if you want to sell an old original video game you know has value, it's either the proper collector channel (slow, super involved, higher pay), the generic resale channels (fast, no effort, lowest amount possible) or ebay (relatively fast, a bit involved but not too much, pay depends on how good you defined the product).

I had some old stuff around that Mr everyday might find (eg pristine original Gameboy Pokémon cartridge without box), and quick resale would give me 10 euro, involved resale would ask me hours and hours of work ebay allowed me to sale for 120+euro spending 1h on the description and picture (to show the scratch etc).

Another case is "oh you have the msi ge77vx4 laotop and you look for the plastic keyboard map in azerty? You can pay a 500e rma if they even allow it or buy the piece for 20e on ebay and fix it yourself"

Ymmv but it has a specific place that no one really have right now


No the interpretation is that the more we could prove it if real, the less we do

Sailors saw mermaids all the time too, I don't think they're all hiding under a rock since we invented the camera


sailors also reported seeing kraken as well, they were eventually proven right with the giant squid.


Exactly, that's the point : if it's true/right, we are now able to prove it with evidence. If it's not, suddently we don't see it anymore.


They reported seeing a lot of other things as well. Rationalizing that as "they were right about big squids existing" is a bit of a stretch.


It wouldn't have the same appeal if they reported seeing "larger than average size squids we didnt know existed". Every story was embellished in a world without pocket cameras. And the further you go back, the more grand the fiction was. Tales of men splitting the sea and walking on water.

It is all very highly entertaining.


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

Search: