jijji

gaslighting people by pretending that encrypting reasoning traces between agent and client is a win for "private AI"... the title should be rewritten as it is plainly conceding -- Closed AI providers use encrypted reasoning traces to hide what the model is doing to come to a conclusion. This is not scientific progress, this is molopoly protectionism. No person using these models wants the reasoning traces hidden from them, it prevents the user from learning how conclusions about a question are derived.

The fact that frontier LLM providers pirated all the data that they used for training, then to go on to encrypt all of the reasoning traces that they use to come up with the conclusions it's really disingenuous, and then have the balls to say distillation is some kind of bad behavior. they are the kings of distillation.

The hiding of this data only brings distrust to their frontier models. I think most people want to understand how something comes to a conclusion they don't want have that part left out on purpose...

it's this kind of behavior that forces people move to to open source models in the end, it's the lack of trust. the frontier model providers treat the end user/customer as a threat or adversary. Fable 5 is notorious for this. a lot of the serious questions you ask the model they won't even respond to you because of the woke guardrails. it wasn't only a couple weeks ago that huggingface had to use glm 5.2 to get the right answers about their security incident because Fable 5 didn't want to answer it.

rust programs are a lot of times faster and orders of magnitude safer than C and C++ counterparts... where is your data to support your assertion that rust programs use more resources than C and C++ programs?

you would think the simulation this would be called via API and get a json response, not be running on the client consuming 1.6 GB of RAM....

something to try that works really well is to put "explain this as if you're talking to a 5th grader" at the end of your request.... it just breaks down the text into more manageable sentences that can be understood by general audiences

it makes you wonder did the developers intentionally add vulnerability exploitation as a feature so they can use it against their competitors or anyone else

taalas is great for llama 3.x 8B models, really bad for one board serving Kimi K3, it seems like you would bottleneck at a few hundred tokens no matter what you do.... spreading the big model against multiple cards seems the only way to get into the 1k+ tok/sec range. Another thing taalas is doing is masking the model weights into the silicon itself, not a flashable firmware, which would increase latency....

thats why i never bought a game in my life and never intend to... if i want to see how a game works, i could download an emulator and run it via the emulator from a public rom under linux. If I am really interested, i could download a 512GB retropie torrent/magnet and then copy it to a microsd card and run 70k games on a raspberry pi 4...

historically, at least for me, games never worked... i spent more time fixing why they didnt work or debugging them in memory to patch them to get the game to actually run then actually playing a game. This is circa 1983-1987 too so maybe we have progressed from there, but obviously not much.

cost of compute, also the less likelyhood of underhanded tactics or vendor lock-in

competition in this space is great, especially with open models/weights. I think the answer is not closed source models. Similar to the Unix versus Linux situation in the 1990's, open source wins out. Yesterdays story about how OpenAI has now began encrypting traffic between model and agent [0], this story brings a breath of fresh air. There is nothing "Open" about hiding the communication between model and agent, especially with software that is running within a trusted environment/network. It needs to be more transparent, not less.

[0] https://www.theregister.com/ai-and-ml/2026/07/15/openai-hide...

s3 is expensive... there are a lot of cheap options. I think I pay $48/month for a linux vps with 8 cpus and 16tb of storage with interserver.net... the same storage on amazon s3 is $377/month lol

what does that have to do with AI companies being forced to exfiltrate data to the US government?

snowden's book, "Permanent Record" talks about the mass surveillance that he released in 2013. it makes no mention of AI companies sending data to the government. unless you can quote the exact language you're talking about in the book, I call BS on that.

secondly, you're claiming that it is the same in the US. there is no laws in the US requiring private companies to exfiltrate data and send it to the government en masse, like there is in China. these laws just don't exist, and if they did exist they would be ruled unconstitutional in violation of the Fourth amendment. As a matter of fact, last week Supreme Court decision in Chatrie v. United States came out on Monday, June 29, 2026 that affirmed that general warrants are a violation of the Fourth amendment, related to sweeping geofence warrants. so what you're talking about just does not exist. Yeah, I'm sure some companies, like Facebook, Google, etc have private agreements with government agencies to release data privately, these rules eventually are a violation of the Fourth amendment without a warrant.

to avoid any of this, I use open models with small providers that are very unlikely to be having private agreements with government agencies specifically to avoid that conflict. it's up to you to protect your own privacy. but to claim that the US government is doing this with AI companies, it's just flat wrong.

Chinese Communist party? it would be funny if China didn't have laws related to companies in China exfiltrating foreign data and giving it to the government for espionage purposes, z.ai is a chinese company, based in the capital Beijing. The likelihood that they work with the government hand in glove is pretty high. I wouldn't run that binary on any machine I was doing anything serious with. If you ran it on a corporate network, I would hope security escorts you out of the building.

or more likely, sending it to the CCP

it all looks suspicious:

  - June 1st 2026: Anthropic files S-1 paperwork with SEC to get ready for IPO

  - June 2nd 2026: Anthropic annouces expanding "Project Glasswing" to let people use their new model to enhance security of existing systems

  - June 9th 2026: Anthropic releases Mythos model

  - June 12th 2026: Model gets export regulations placed on it by US Gov

  - June 26th 2026: US gov announces they will let some companies use new model

  - August 2026: Anthropic goes IPO

The timing of all of this just seems to be a play to pump the stock. The reality is that in six months GLM-5.3 will be released open source with comparable functionality to their Mythos model. They are trying to cash in before that happens.

I would not be surprised if the US government, the people pulling the strings who actually put the export announcements onto Anthropic, actually have purchased stock in the company to artificially pump up the stock, I would bet money on it.

most large companies have a 2-year limit on contractor employment so what they tend to do is they'll hire the same guy through a different contractor with another two-year agreement..... that's to get around the situation where if someone is working as a contractor for more than 2 years they can legally claim that they're actually an employee....see Vizcaino v. Microsoft Corp., 120 F.3d 1006 (9th Cir. 1997) [0]...

this is just a guess by the way but it seems like a plausible one, as I've seen it happen in Fortune 500 a lot, where the same guy comes back through a different vendor 2 years later if he was really good and they needed him to come back....

[0] https://law.justia.com/cases/federal/appellate-courts/F3/120...

glm-5.2 is available for $20/month on ollama.com and is IMHO more functional than the $200/month claude max subscription. you can even use the same claude harness [0]. You get about 20x more token usage at 10x less the price.

[0] https://ollama.com/library/glm-5.2

if a "cell" is different from a "container", from the articles context, please explain how... This could be some special FPGA board too referred to again as a "cell", but they are extremely vauge about the definition.

I agree, confusing and it seems like they are using the coined word "cell" to describe "container", but they really should say that instead of making up new words.

have you you checked out "esp32-s3" which costs $7.12 and has wifi and microsd installed on it [0]. Also esp32-cam is another board with similar specifications.

[0] https://www.seeedstudio.com/XIAO-ESP32S3-p-5627.html

sounds really cool if it was coming out of anywhere except China, which has laws to exfiltrate your data and send it back to the government for espionage purposes [0].

[0] https://www.fbi.gov/investigate/counterintelligence/the-chin...

They are a few years late to the party

Using the demo and typing in "A sign that says xxxx" where xxxx is any text, it gets it wrong almost 100% of the time.

I think you mean "StarOffice" which later forked into OpenOffice, then LibreOffice.

I think Microsoft stopped being the "darling" in 1994 when they got sued by Stacker and had to pay $120 million for stealing their source code and using it in their own product.

The situation you reference is related to a specific investigation by US congress requesting documents about potentially illegal censorship actions by EU officials from a specific company (microsoft). The difference is that the laws in china are broadly defined to include giving all intellectual property of anyone back to the government with no oversight, for the purposes of espionage.

The former relates to a specific investigation about potential criminal activity, the latter relates to broad illegal activity committed by the government itself unrelated to any specific case.

The US has no laws on the books forcing companies to wantonly give intellectual property and other espionage level material back to the government. If they did, no one would use cloud providers.

To avoid this, you can run your own hosted machine in a colocation facility, because in the US, people do have reduced rights when their data is controlled by a third party versus being controlled by themselves. Its the same as if the data was in your house, they would need a search warrant to obtain it, but when its at a Azure or AWS datacenter not controlled by you, your privacy rights are reduced by doing this.

there's laws on the books in China that says that every company operating in China must aid and abet the Chinese government in espionage against the rest of the world. given those facts, I find it deeply troubling to be using anything coming out of China, especially a program that runs in the context of a Linux terminal on a machine that might have something important on it. I'd argue it's a back door waiting to happen, if not sooner than obviously later.

I just can't get past the deepseek-CCP connection... as good as it might be I'd wonder when your machine gets backdoored by the CCP or at least your data gets stolen

I've seen good results with opencode connected to glm 5.1 on ollama cloud... for $20 a month you get similar performance that you get with opus 4.7