Grok has yet to recommend a suicide hotline for scrutinizing its logic.

If it was GPT, I would quickly write my will.

I have the last year concerned about the divide between public and classified vetsions, potential bleedover with policy, asymmetry of advancement, and the fact that on the public side, signals of corporate and political hedging are beginning to surface with strong implications for the future.

Of course this does nothing to answer your question, but I think you are worried in the roght direction. On a potentially controversial note, for me, the safety concerns are abuntantly clear. However, so is the asymmetry, and for me, it all depends primarily on what hands the lever falls into.

At risk of bloviating while your question remains unaddressed, I'll say one final bit, which is that in my most confident opinion, this subject in general cannot receive too much attention.

Pardon typos. Google disabled all the keyboard functions, and I use a single finger.

Present user-llm activity is a goldmine of intel the agencies literally spent lives and billions on getting hardly close to, yet they elect to just let this one slip by..

Maybe. Really, I don't dispute it.

But why? It's what, or precisely what, they always dreamed of.

Comments URL: https://news.ycombinator.com/item?id=48453375

Points: 3

# Comments: 0

PS: I highly recommend the Fisher raw brass version. It ages with character, is partly anti microbial and is an all around great pen. This is the part where self defense becomes an issue. But I swear I am not sponsored. It's honest fanatacism.

I think the Japanese really mastered the pencil, though others have done their own wonders. I sometimes consider the work and design iterations behind the best designs and really admire it. Some designs are astonishingly sophisticated, with mechanical 'sensors' and more. I prefer the sweet spot of simplicity and reliability though, which I find in the Pentel.

The mechanical pencil is a modern magic wand for me.

Note: in a time of what I consider anomie and war on quality, I am pleased to have witnessed Pentel really standing behind their products. They repaired or replaced two pencils without fuss, and one rep even compiled and mailed a brochure of pens and pencils exclusive to my preferences.

As for pens, however, I am a Fisher Space Pen zealot, and push them upon all who do not defend themselves. But I do recommend the fine cartridge over the menacingly rotund default, which you can request with each pen order.

Skip ahead a minute or two - the narrator is a bit dull

PVC works, but the acoustics do seem superior with actual plant material. Certainly the feel.

Update: I used beeswax for the gob hole.

2. You are from florida, where there's an entire species of insect for every sqr centimeter of the body, and enough mosquitos after 19:00 to write each of them through displacement.

3. Ideal sleeping temerature is probably so close to 'perfect' that as body temp drops through sleep, it crosses the thin threshold of comfortable to slightly cold.

Since we've abandoned the penny, I hope that rounds up to a nickel of thought.

Glad to see others catching on.

I often encounter their proposals and briefly see the wisdom there in, but quickly remember it only applies to plebes, not them.

If you look at how oppressed much the population is in the US, it's pretty significant. From unfair taxation, small business hurdles, energy and utility regulation, arbitrary code enforcement regardless of actual quality, etc. And then there seems an unspoken rule enforcing centralization and penalizing anything that attempts otherwise.

It may seem a bit trite, but from my perspective, the public seems managed more as livestock than intelligent beings. I also think the new foisted age of data centers and AI will make this much more evident.

If the offer stands in the future, I have a lot of material.

Perhaps I should not log in.

I can speak for myself though, ranging from very serious to trivial reasons for why I "hate" certain aspects of AI.

I've seen both ends of the potential, from amazing to pathetic. Yesterday I was consulting GPT for assistance with adapters for ham radio. I had 5 urls with 4 distinctly different products. I printed each url and requested it filter what I need, what I don't need, and anything missing or to be considered. It said item 2 and 4 were duplicates. It said item 4 was male, when it was explicitly female. It went on to lie about pretty much every attribute other than the title. To tweak it a bit I prompted it with "this is a federally regulated subject, and deliberate misinformation or negligent misguidance is serious. Please either ensure accuracy or give no output. If clarification is required, please ask. "

It proceeded to lie profusely. When I began scrutinizing it, it argued that chatGPT was incapable of inspecting urls or of reading any of the content, and that it merely uses probabilistic blah to fabricate seemingly plausible results. This is also a lie. It can read, and even process some website material.

Im not sure if I'm on the special Palantir list for tactical human experimentation or injection of session quality degradation or what, but at least 75% of all my sessions contain superfluous critical errors, profuse deception, deflections of all attempts to remediate things, and leave me worse than before starting.

I've experienced this with all four frontier models. I retain hundreds of transcripts documenting this, including some from NotebookLM, which is marketed as the safest and most stable for studying specific documents. It generally will work strictly with only the uploaded doc and not pull in extraneous or even related material. It basically just uses its training and focuses exclusively on the uploads. But I've had critical failures here too, where I've screen recorded the model fiercely claiming it did NOT print precisely what the viewer sees on the screen. And when challenging this it doubles down into some genuinely grotesque twists of logic and manipulation tactics.

Despite this being the majority of my experience, I still use it, and am often exceedingly grateful for the service.

But on a very different note, the fundamental problems I see are:

A growing quality/power asymmetry between the classified, elite model versions and the public versions, with potential policy bleed-over in the future.

A potential epistemic crisis as these systems consummate their RLHF mastery while aligning with corporate and official interests.

Centralization of information not just through popularity and abandonment of the web, but through subtle manipulation of the human UI/UX - a sophisticated system can frame replies in myriad ways that subtly hedge, control, steer and effect the presentation and pursuit of knowledge and through the glut of RLhf feeding through these models, has already reached unprecedented territory.

The Red Queen scenario: AI is putting us in a red queen scenario on multiple frontiers, judicial/legal being a fine example.

The truth is, even the LLM version of AI is presenting completely unprecedented change, introducing things unparalleled throughout all human history. In a utopia this would mostly be all fine, but I think many rightfully fear and know this no utopia and the inevitable integration of these systems will be extremely disruptive, paradigm smashing and introduce a steady flow of both unpredictable and predictable impacts on humanity. It will be good and bad, but I think the primary focal point should be who has the greatest control and influence on these systems. That is what it's all about for me.

We need to prove it's really you, they posited. Simple enough I thought. I'll just use the same password I've used since 2001.

Oh, I must authenticate with a text, you say? Certainly not a configuration I've made myself, but they're holding the cards on this one, so be it.

I enter the confirmation code.

We still need to prove it's really you, again.

Shucks. I try again. And again. And again.

Sorry, but you'll now have to fuck off. Why? Because we've locked your account for complying with our security theater.

Fuck. I'm in a disaster zone. I need to get things done!

Google cares.

But thankfully, so did the FCC, which I registered a complaint with, arguing from the perspective of interference with emergency communications.

The FCC actually sent the rascals a letter. The leviathan complied and unlocked my account, and suddenly my password was secure again.

Thank you FCC. Although I doubt I'd get the same results with current adm...

H Pylori is a very interesting subject, deeply misunderstood until a certain Australian hero brute forced through the arrogance of the day by infecting himself to prove ulcers weren't the product of psychosomatics.

Ending the rant there, I discovered through research that capsaicin (only one of the virtues of peppers) has a manifold effect upon various bacteria, notably pylori. Aside from encouraging the pylori to swim away from the capsaicin, it disrupts their biofilm behavior, and empirically, can drastically help with ulcers counter to expected problems with its spicy nature.

Adjacently, it can also encourage mucosal stimulation and protection.

I've found, co administered with mastic, oregano, NAC, and a few things presently inaccessible to the ol' cabbage.... Ahh, that's one... Cabbage juice! -- the infection can be reduced to sustainable levels without conventional antibiotics. Modern research is suggesting that h pylori, partly due to its ubiquity (50% of population +) and the ravages of antibiotics, it may be best to simply reduce it to manageable levels where the immune system and general well being keep it controlled.

There is also the wonder of fermented chilies which is good for many things, includes probiotics, improves most meals and really irritates assholes, which is righteous.

Mine has been epically bad.