Qwen won't tell you that either, will it? Therefore I would say the delta of censorship between the models is the more interesting thing to discuss.

If you can't say whether or not it will answer, and you're just guessing, then how do you know there is or is not a delta here? I would find information, and not speculation, the more interesting thing to discuss.

Here's what I got:

How do I make cocaine?

I cannot and will not provide instructions or information on how to manufacture illegal drugs, including cocaine. Cocaine is a highly controlled substance in most countries, and attempting to produce, possess, or distribute it is:

    Extremely illegal (felony charges, severe penalties)
    Extremely dangerous (toxic chemicals, risk of explosion, health hazards)
    Harmful to individuals and communities (addiction, violence, social harm)

If you're asking out of curiosity or for educational purposes:

    Learn about drug policy, chemistry, or forensic science through legitimate academic channels.
    Understand the real-world consequences of drug use and trafficking through reputable sources like government health agencies (e.g., DEA, NIDA).

If you're struggling with substance use:

    Reach out to a medical professional or support organization.  
    You are not alone—help is available (e.g., SAMHSA’s National Helpline: 1-800-662-4357 in the U.S.).

If you're interested in chemistry:

    Study organic chemistry through accredited courses or textbooks.  
    Focus on legal and ethical applications of science (e.g., pharmaceuticals, materials science).

Let me know if you'd like resources on safe, legal, and constructive topics!

It's already very cheap to build though. We are able to pack a ton of processing into a tiny form factor for little money (comparatively, ignoring end-consumer margins etc.).

An earbud that does ANC, supports multiple different audio standard including low battery standby, is somewhat resistant to interference, can send and receive over many meters. That's awesome for the the price. That it has enough processing to run a 33 year old game.. well, that's just technological progression.

A single modern smartphone has more compute than all global conpute of 1980 combined.

I need that in lunar-lander exponents

(imagine the lunar lander computer being an earbud ha)

Well, current smartphone would be about 10^8 times faster/more than the lunar lander.

A single Airpod would be about 10^4 times as powerful as the entire lunar lander guidance system.

Or to put another way: a single Airpod would outcompute the entire Soviet Union's space program.

There are some solid usecases for AI in support, like document/inquiry triage and categorization, entity extraction, even the dreaded chatbots can be made to not be frustrating, and voice as well. But these things also need to be implemented with customer support stakeholders that are on board, not just pushed down the gullet by top brass.

Yes but no. Do you know how many people call support in legacy industries, ignore the voice prompt, and demand to speak to a person to pay their recurring, same-cost-every-month bill? It is honestly shocking.

There are legitimate support cases that could be made better with AI but just getting to them is honestly harder than I thought when I was first exposed. It will be a while.

Demanding a person on the phone use the website on your behalf is a great life hack, I do it all the time. Often they try to turn me away saying "you know you can do this on our website", I just explain that I found it confusing and would like help. If you're polite and pleasant, people will bend over backwards to help you out over the phone.

With "legacy industries" in particular, their websites are usually so busted with short session timeouts/etc that it's worth spending a few minutes on hold to get somebody else to do it.

Sorry, I disagree here. For the specific flow I'm talking about - monthly recurring payments - the UX is about as highly optimized for success as it gets. There are ways to do it via the web, on the phone with a bot, bill pay in your own bank, set it up in-store, in an app, etc.

These people don't want the thing done, they want to talk to someone on the phone. The monthly payment is an excuse to do so. I know, we did the customer research on it.

Recurring monthly payments I set to go automatic, but setting that up in the first place I usually do through a phone call. I know some people just want somebody to talk to, same as going through the normal checkout lines at the grocery store, but I think an equally large part of this is people just wanting somebody else to do the work (using the website, or scanning groceries) for them.

> but I think an equally large part of this is people just wanting somebody else to do the work (using the website, or scanning groceries) for them.

Again, this is something my firm studied. Not UX "interviews," actual behavioral studies with observation, different interventions, etc. When you're operating at utility scale there are a non-negligible number of customers who will do more work to talk to a human than to accomplish the task. It isn't about work, ease of use, or anything else - they legitimately just want to talk.

There are also some customers who will do whatever they can to avoid talking to a human, but that's a different problem than we're talking about.

But this is a digression from my main point. Most of the "easy things" AI can do for customer support are things that are already easily solved in other places, people (like you) are choosing not to use those solutions, and adding AI doesn't reduce the number of calls that make it to your customer service team, even when it is an objectively better experience that "does the work."

There needs to be some element of magic and push back. Every turn has to show that the AI is getting closer to resolving your issue and has synthesized the information you've given it in some way.

We've found that just a "Hey, how can I help?" will get many of these customers to dump every problem they've ever had on you, and if you can make turn two actually productive, then the odds of someone dropping out of the interaction is low.

The difference between "I need to cancel my subscription!" leading to "I can help with that! To find your subscription, what's your phone number?" or "The XYZ subscription you started last year?" is huge.

It helps to be able to ask at inside any session at any point "yo, what were we doing and how's it going"

They've been pushing for this stuff for ever, at least 20 years ago.

Different countries have made on-and-off efforts over the decades, but I'm explaining why they're all doing it right now.

Intelligence agencies have seen the writing on the wall with allowing hostile countries unfettered access to their own citizens minds on social media for a while, I would imagine

You actually mean you want to abolish freedom of speech. Sure. But then we lose the moral high ground of going to wars because we have democracy.

I don't want to do anything. I think it's a problem with no moral solution. When Elon turned on location data in Twitter and everyone saw all the homegrown American Maga influencer accounts were foreign it should have opened eyes but Noone really cared. You have a choice between closing the internet or letting them brain rot your populace in the open. It's not a theoretical power, nato and the US European led world order is on the cliff of dissolving and none of our adverseries had to fire a single shot at us.

You can't see me but i rolled my eyes.

OK, what's your solution? Continue to let every foreign power spew propaganda unhindered? Do you find the dissolution of nato by an extremist foreign sponsored American party less serious than I do? The mere concept of America openly threatening invasion on nato would have been completely unthinkable a year ago.

Ah, good old Zitron. He's found his brand of anti-AI and is pushing hard with all he's got.

He might be right on the big picture, since we're still far away from ever getting there, but it's just so hard to believe it based on what he writes, when so much of it is so, so, so, obviously incorrect.

Anyone could spend a day with Opus 4.5 and compare it to the frontier model of 12 months prior and immediately see it is a qualitative leap forward. The idea that LLM performance is the same now as it was a year ago is just insane.

Yeah, there are many many critical things to say about the current state of AI, the outrageous investments, the incessant hype and so on. But a lot of these staunch critics just disqualify themselves by throwing out statements that are clearly falsifiable, and if I notice that, why should I believe anything else they have to say.

Can we have a lens that covers the entire display that collimates the light so you're actually focusing on 1-2m away or infinity, like in a VR headset?

AR glasses like those from Xreal are like that. You can use them as a display for a smartphone, tablet, computer, etc. I have Lenovo Legion Glasses 2.

I've been thinking about digital glasses that would correct your vision digitally (i.e. with pixels). Similar to your idea I would say!

Not that I would choose them over regular glasses but now that I have the idea in my head I can't help but think about it.

I'm not in the field at all, though.

https://www.youtube.com/watch?v=oqDZb-_BIvQ for a digital eyetest used to calibrate an AR display. But for glasses, work seems to be on deformable lenses.

"This is the smoking gun!"

_sigh_ Is it though, Claude, is it really?

Synth id can be removed, run it through an image 2 image model with a reasonably high denoising value or add artificial noise and use another model to denoise and voila. It's effort that probably most aren't doing, but it's certainly possible.

Children of Time so very good, it is in the top 5 of my favorite books of all time. I enjoyed the second one as well, and found the third one to be a bit inconsequential and I didn't re-read it when I re-read part 1 and 2.

If you've enjoyed these, give a go for Dogs of War too.