If your goal is to understand the quality of the espresso shot, rather than experience a high quality espresso shot, letting it cool off provides a useful data point.
> Schiendelman sent you a message on hacker news clarifying what its showing. It's you asking about the plant she told you about, it finding that text message, summarizing, and also proving with the text that it's accurate.
In the simplest possible terms: this is total bullshit security theatre. At no point has there ever been a bomb or even a bomb threat carried out via usb device names. There is absolutely no reason to even look at the names of Bluetooth devices on a flight.
I’m certainly not joking. Google when it started it wasn’t as evil as now, but the bigger it gets the more evil it becomes, who knows what kagi will turn into if they got as big as google. But again on principle, can you use google search in the library without an account? Yes. Can you use kagi in the library without an account? No. So whenever and whatever you do, your queries are logged and tracked back to you, only waiting for xyz to be pulled out.
Google still lets you do some things without logging in but that doesn’t mean that they don’t build profiles or try to link them with other activity sources. Most of their revenue comes from advertisers paying for targeting.
The paranoia but also the naïveté of internet tracking is a bit of a rare combination I think. Shadow profiles have been a thing for 15 years or more and they have only gotten more sophisticated. Browser fingerprints are surprisingly unique. Unless this guy is rotating machines / vpns / using qubes / etc Google can very likely pinpoint within a degree of certainly which searches are theirs.
It isn’t paranoia when there are clear facts in here, the first fact is your queries are linked to your account, period. Second fact, you don’t control the server in any way or shape. Third, your identity is linked through your payment information, aka full legal details, not just through an IP. Comparing between all that and browsers fingerprinting is muddying the water, you can use searxng in the middle and anonymize the query, with no profiling cookies etc., adding few extra steps and it’s private for most people unless you are a person of interest, in kagi, this doesn’t exist! Even the anonymizer is ran by the same company, so much trust!
I am not paid to design kagi architecture nor I know the internals, but let’s say I can host that mentioned anonymizer myself (say in Canary Islands), and it pulls the queries from kagi who you have paid for by monero, then the company knows nothing about the user, no profiling, no tracking, nothing, that’s a great starting point.
If this doesn’t exist, using something like searxng is far better (privacy wise), not just as mentioned on how it’s more anonymous, but also it gives you the ability to blend in, rather than looking like a sore thumb in the logs.
The cookies used by the site appear to almost all be from Cloudflare, GitHub, YouTube: service providers that are not at all interested in enabling a cookie-free web site, and chose this route as malicious compliance. There's also one cookie to store a language preference that could be handled through HTTP headers instead.
The title is misleading and not in the article. This change is for business/enterprise accounts. Also, these are still credit based. The change is that credits now operate on tokens like the API rather than on messages as they used to.
> Customers on existing Plus, Pro and Enterprise/Edu plans should continue to use the legacy rate card. We’ll migrate you to the new rates in the upcoming weeks.
Nope, they buried the lead a bit but this is coming for _all_ users, even pro/plus subscription plans. So you get chatgpt pro/plus benefits, and then effectively $20/$200 in credits for codex
First of all, there's no dollar amount tied to how many credits you get for a subscription.
Second, if you look at the prices for bundles of _extra_ credits and then do some math on the Codex rate card, you'll see that there's no way they would work out to be the same or similar.
> First of all, there's no dollar amount tied to how many credits you get for a subscription.
I don't understand what you mean here; their official comms is:
Customers on existing Plus, Pro and Enterprise/Edu plans should continue to use the legacy rate card. We’ll migrate you to the new rates in the upcoming weeks.
To me, anyway, that means that GP was exactly right - they'll give the $20 subscriptions $20 worth of credits, and the $200 dollars subscriptions $200 worth of credits. That is what the "New Rates" are!
I think it would be more rational to discount a subscription (standard is about 10% in most industries) vs PAYG and agree in principal with your assertion - they haven't specified what the discount is on credits bought in a subscription plan - but there is no indication that they are going to continue allowing thousands of dollars of credits on a $200/m plan.
My guess would be a 10% (or similar) discount if you buy a subscription.
The TurboQuant paper is from April 2025. I’m sure the major labs knew about it on, or even before, the day it published. Any impact it had would have been a year ago. Yet I keep seeing these posts and discuss completely ignoring this.
Can we please start talking about this in that context? We already know what TurboQuant will do to DRAM demand. We already know what it will do to context windows. There is no need to speculate. There is no need to panic sell stocks.
It could also be that masters degrees concentrate in fields with lower compensation. Teachers are in high demand, but yet they still tend to have something beyond an undergrad.
reply