Because they are preparing Gemini 3.9 Flash

The real singularity is when new models are released faster than we can change the model dropdown dialog.

When I was in the US as a student (F1 Visa) that is why I decided to go back home instead of trying to optain an H1B. It was just too risky for too long. I did not want to live a life where I had to expect to leave the next day. Besides, it was right at the beginning when Trump won the first time.

When you open it, it becomes a tablet not a phone. And therein lays your answer.

That might explain my naivety here, I've never really gotten into using tablets for anything.

My mother uses it for everything. The small phone screen is hard for her eyes, she doesn't really need a laptop and the tablet is just right. She still carries a phone with her but I can imagine if she replaced both tablet and phone with a single foldable she'd be happy. On the other hand, she'd want it to last 5+ years and she'd never spend €1800+ on it.

And I've never gotten into using my phone for anything outside of calls messages and navigation. The web is the worst on tiny vertical screens.

I am a tablet user but crease is a non starter. However I know people who like foldables in different ways, like financial stuff, they don't care

For most, if not all users, the crease is not something one notices during regular use. You just forget it's there when you look at content.

That said, with the Pixel's enormous crease you feel it if you touch the middle of the screen.

Depends on your use. If you draw or want high quality display with no weird reflections from your tablet then it's a non starter.

Yep! Having just got the Pixel Fold 10. I've been wanting a tablet for casual browsing and video watching but I knew I wouldn't use it because it's not my phone. The apps data would be different logins are annoying. If I'm getting another device out I'll just get my laptop. This thing is great though. There can't be a more seamless transition. The multitasking is pretty good and with coding agents it's actually turned out fairly production to have two apps side by side.

a tablet with a crease right in the middle of a screen (FTFY)

You really don't notice it while you're using a folding phone.

someone said "tablet". When I use a tablet I personally don't just notice it, I literally can't stand any uneven surface under stylus or uneven reflections.

So the unfolded state of the folded phone is not a "tablet" to me. Maybe a tablet with a glaring defect... definitely not a tablet I could use as an actual tablet.

I can't tell if you've actually tried a finding phone or not, but that would help understand and place your comments better.

I have, I have a friend with a folding phone

> I use it but feel uncomfortable, that it has large attack surface and LLMs will find exploits in it

Doesn't this apply to any application you use? How would it be different with plain wireguard?

> How would it be different with plain wireguard?

Seriously ?

You do realise that of all the security tools on the planet, plain wireguard most likely has the smallest attack surface of them all, right ?

The problem here is as the other poster said. Tailscale is a security tool and yet the guys at Tailscale seem to be insistent on dumping everything INCLUDING the kitchen sink into it as a "feature".

That sort of attitude is not going to end well. You end up with a large bloated code base, which equals large attack surface.

I am talking especially about the LLM part.

Also a kinder tone in your comments would be more appreciated.

No. If software is small enough, it can be proven.

There is formal verification of the protocol and aspects of code:

https://www.wireguard.com/formal-verification/

The code is small enough that can be reviewed.

That makes sense. Thanks!

Why do you say lazy? maybe they are ok with people seeing it is claude?

Why do people run and install these agents locally? No container nothing. I am running Opencode in WSL2 with the windows mounts disabled.

So the idea is that these should be treated as programs in an extremely low trust environment, akin to running malware in a VM?

yes, this is basically experimental tech, if used with open source harnesses. if used with proprietary harnesses, treat as actual malware.

Technically open source harness like opencode is more "malwarey" than claude code for example because its default permissions are very open.

I disagree with this, Claude code does many malware like things and has its source obfuscation. Plan mode doesn’t even work.

I never thought about it in that extreme, but just today the google results ai gave me conflicting information from one search to another. Maybe I should start.

P.S. the conflicting information was Keith Richard's age. One search said he was 37 Dec 18 1981, the other said he was 37 in 1980.

I treat all proprietary software as malware, though of course the risk surface varies

Yes, I think if agents as interns with enough smarts to be dangerous.

If the agent you are running really wanted to it could easily find a way to mount the windows folders and read them all. WSL isn't a security boundary, you are only barely more protected than people running grok in their home directory.

it is not about "security" but it is more about confinement. And how could it do that without sudo rights?

Still, you are barely more protected than someone running this as non root at home. By your same logic you could “confine” the agent to any arbitrary directory on your filesystem and say it’s sufficient.

Yes, that is exactly the point I am making. The agent / cli is confined within the WSL2 environment. How is it "barley" protected? Neither you nor the other guy have yet answered this question.

I never claimed WSL2 is a security sandbox. I am saying running it in a container or WSL2 allows you to severly limit the blast radius. I am not expecting the agent to be malicious, but I am expecting it to do unexpected things.

That’s a fair approach, thanks for elaborating. I think the main point of contention is there is an expectation this software should be more reliable, that doesn’t take away from your approach, running close sourced software like a harness (or operating system) is asking for problems.

Same reason one uses VSCode locally

I use Openwebui

The context window size is also very small if you use Gemini in the app. It starts forget quite fast. In my opinion Gemini on app is useless additionally to the guardrails.

It is the same reason I decided against getting a motorcycle (I also decided against getting a PPL as well).

Yesterday I needed a tool for a specific task. In 1 hour I created a working tool with antigravity. Is it something I'd publish or sell or offer to others? Probably not, but I found that quite impressive. It is an extension of the idea "now I can program and create everything."

Besides I remember what kind of code we had when we first started coding with AI and now with all these coding agents etc. it has become quite impressive.

However, at the end it is still a tool and needs to be used accordingly.

You can create anything you need as long as what you need is a disposable script, a scotch-taped together single page app, or a complex problem and you have thousands of dollars to throw at tokens.

I've been playing with local models for some time, and I've been pleasantly surprised of late. A meager rtx 5080 with 16gb can give pretty good results now. The ecosystem is also improving pretty quickly.

I have a feeling at some point we will have a "Windows 95" moment (when computing really became personal for the masses) in AI, and things will significantly change shape again.

What local model do you recommend these days? I’ve got a 4090, mostly sitting idle.

The answer to which ai model, in mid 2026, is always qwen. Depending on your ram, it’s qwen3.5-9b, qwen3.6-35b-a3 in a 3 or 4 bit quant, or qwen3.6-27b. I’m told a bigger model quantized is better than a smaller model unquantized. In 16Gb vram on 10 year old hardware i can run a 3bit quant of qwen3.6-35b-a3 at ~30tokens/sec, and it can do a lot.

qwen 3.5 with 9b is being a pretty decent workhorse for me, even with context around 4k.