Locking this to X accounts destroys nearly all of the value for me, because the data is limited to only setups from people who are still using X. Given how overtly authoritarian and hateful the general rhetoric is on that site, this narrows it down to only a specific kind of person's setup. Given the context of what X is, one can largely assume that X account holders are biased, at least to the degree that they can tolerate the toxic rhetoric elevated by that site and it's owner.
All of the smartest, most effective people I know don't have X accounts anymore on ethical grounds, and don't miss it. If a data source specifically excludes the kinds of people I respect most, that data source is all but useless to me.
I don't think ignoring is the right choice. This site has power that must be challenged.
One of the biggest issues I see is that governments still use X as a primary form of communication with the public. We need to push back on this with lawsuits.
There is absolutely no reason I should have to agree to a private company's TOS to receive communications from the government. It is a violation of my rights that refusal to agree to the X TOS would cut me off from public information.
It's fine if they post to X, but we need to demand laws that the government isn't allowed to have special private clubs where the general public isn't allowed without giving up some amount of their speech rights.
>One of the biggest issues I see is that governments still use X as a primary form of communication with the public. We need to push back on this with lawsuits.
This.
I’m in the northeast and I’d love to not have to go search x/facebook etc to find out it’s a snow emergency (which city dept will post this?). My city has a website, use it.
Honestly, with apps like Immersed, custom two-hand keyboards, and trackball mice, it is already feasible to lie comatose in bed with a VR headset strapped to your face and get through an entire day of software development.
This is my first time hearing about RTK, and yikes! The benchmarks mean nothing; this thing is actively dangerous to use.
If my agent runs a shell command, show my agent the output of that shell command. I don't have a problem with automatically pruning or paginating large outputs, as long as the agent still has some form of access to the original output (e.g., by searching).
But removing verbosity from an 'ls' command? That's ridiculous! If my agent runs `ls -la`, do not drop the owner and date from the output, because THAT'S AN EXPECTED PART OF THE OUTPUT. Nothing should be trying to predict the agent's intent and interfering, unless you want your agent to get confused and fail.
These LLMs were trained on predictable shell behavior, and RTK deliberately subverts the model's expectations. There's no way that isn't degrading the model's capability.
For real, I have like 200 kilos of cables. I could very likely reduce that down to something less severe, but that involves sorting and inventorying 200 kilos of cables.
Just begin. You don't need an inventory. Just start by dividing them into piles on the floor, organized by role.
The biggest piles can get further subdivided into more-specialized roles.
There will be piles that consist of memorable cables that have interfaces you haven't used in years. Segregate these unicorns; they'll be handled with special care later on.
After that, you've got a broad selection of current-vintage stuff that is in active use -- all laid out before you. So just spend a few minutes to pick out a few nice examples from each remaining pile. A few cat5-ish cables of various lengths. Some AV interconnects of different descriptions. A few universal power cables: IEC, figure-8, and Mickey Mouse. Pick out the very best of the USB cables, and of the SATA cables, and whatever else you're making use of in 2026.
Wrap each of those selections up neatly and put them in appropriately-small boxes. Label each box by role. Put each of these boxes away somewhere that is both out of the way, and able to be quickly returned to later.
That was easy.
The remaining piles on the floor are just scrap. Make a plan to scrap them and package them appropriately. Put that package near the front door.
---
Now we get to the really fun part. All those precious historic artifacts that were carefully isolated before? It's time to sort those.
Who knows when a 24-pin dot matrix printer could turn up that you need a DB-25 to Centronics cable for testing with, or you might find a use for the 15-pin PC joystick splitter? That PSX controller extension was hard to find even when it was new.
It's all very important stuff with very specific roles. It's not like any of these things are being made in factories every day anymore. And it all has history.
So take a photo of this amazing collection of vintage cabling, and then package it all with the reverence and care that it deserves. After that, simply put that package near the front door with all of the other scrap cabling.
To be fair, this example doesn't work because those chip bags are two physically different objects that one can consume.
Sony is trying to argue that because they are selling digital copies, that consumers cannot be granted rights to their copy because they can't transfer the intellectual property rights, which is complete nonsense. I have a 1999 DVD copy of The Matrix that I have owned for 27 years, and short of the physical disc-rot, I will be able to play that DVD until the day I die, and Warner Bros can't do anything about it unless I try reselling new copies, or claiming authorship because, despite owning the disc, I do not own the rights to the movie...
A download isn't any different just because it lacks a physical form.
Transparently, what we have been discussing is to offer our keystone products (search, assistant, translate, email, ...) as standalone items. Then we will offer a all-in-one Kagi bundle that has everything under one payment.
The reason for this boils down to the following observations:
- Cost of operation at scale varies wildly between these, making it complicated to operate and manage cash flow
- Complexity for the consumer gets unruly as arbitrary things are limited to our fixed brackets of standard/pro/ultimate, we'll need some crazy matrix to explain what is included in each one
- Apprehension in our customer base about paying for things they don't want, chiefly AI related features. Some people just want search and don't want to pay for anything else, even if its included for free.
Itemizing our ecosystem addresses each of these by letting people pay for what they want, and ignore what they don't. If you want Kagi mail, you just get Kagi mail - there is no gymnastics of "I want Kagi mail, but I need extra storage so I need ultimate, but I don't need Assistant stuff, ...". From a business POV it will cleanly fund each initiative without everything competing for the same $10.
This is still on the whiteboard and not set in stone - so don't take what I'm saying as gospel - but it is generally where our heads are at in terms of managing the complexity of our offerings, cost of operation, and consumer piece of mind. It's a very difficult problem to navigate!
Edit: I mention Kagi Mail here, some people may not know but it is in a closed-ish beta at the moment - we hope to bring it to market by EOY!
I have to imagine assistant (AI usage) will be next, and swap to pay as you go. And search won't get cheaper because I imagine they've made the mistake of treating the unused AI credit as a subsidy for other services
Our $10 search sub price is solely based off of the cost of operating search and has remained unchanged for years (est. 2023), before we offered any LLM-based features as part of our plans.
We added AI features to the existing plans at no extra charge. The only thing we preempted was our Ultimate plan (which we initially offered with literally zero extra perks!), anticipating that LLMs would become more capable and popular and require a higher margin to operate at scale.
No.
reply