meltyness

> Every WaPo subscription comes with free access to "headspace" meditation training, in case you're "feeling overwhelmed by it all".

(1) Frustrating, but surmountable for the time being, what with OPT, LLaMA. Market structure at this early stage may lead to something odd like Springer, Elsevier, Oxford, or Wiley getting in the mix since the weight of Google has been crushing publishing everything for the past few decades. And offerings in this space are paltry, see Microsoft's BioGPT for example (not regarding it's efficacy, since I'm not even qualified to comment but rather that it's a tiny team, and isn't really even a product).

(2) This isn't unprecedented, see fairness doctrine[1].

(3) The beneficiaries of which are the ones that have the largest most brittle tech stacks, short-term.

[1] https://en.wikipedia.org/wiki/FCC_fairness_doctrine

I have some success if in each subsequent prompt I include a reminder about which letters ChatGPT has guessed, and accumulate and present everything a bit like a game of hangman.

That should be prompted, but it's lazy about sorting through it's own outputs like this.

Say you explain the rules and it guesses, "pecks",

> P E _ _ _

> The word contains a C

> The word doesn't contain K, S

Re-promoting and accumulating those details like this, ChatGPT 3.5 managed to solve yesterday's wordle successfully, but it took 10 tries to land on this prompting method.

Maybe more parallel GPTs instructed like, "find 5-10 key facets of reasoning about the problem presented", and then "extract those facets from the conversation, organize them into a table, and use them to supplement decision-making" might help prevent the needed parameters from blowing up too much.

Performing this would require a lot of compute resources.

Perhaps any tasks that require a great deal of dynamic memory/comparison would be expected to struggle in this architecture.

Since "write down all possible permutations of these letters" is like n! "cross reference the permutations with a dictionary/definition if you feel like it" must be performed m*n times, maybe other such n! tasks are doomed until more parameters are feasible.

It almost certainly could produce a bash script to do that based on what I've seen.

This would seem to address the form of the censorship employed, whose outcomes have been controversial.[0]

[0] https://news.ycombinator.com/item?id=34691246

Law is artificial intelligence, so technically we're already in one of those scenarios in most of the world.

Law records what most people have expressed is desirable behavior in society.

Presumably the extremely adjacent concept of knowledge which may be encoded into DNA may also be thought of in this way, so that even under "the law of the jungle".

DNA records what forms can survive and replicate in a given environment.

... "AI takeover" is already the status quo.

Shame is the word.

This may be in support of the 50bp hike the media was stuffing into Powell's mouth all last week after his statements before the Senate, find out where else powerful people have built crab traps.

Maybe ideologically, but if you're still up on the inflation hedge narrative, I suppose you just let it happen.

SaaSs are scammy especially when they involve:

- Repackaging an existing codebase or business model

- ... as a black box

- ... attempting to rent it out

- ... with worse data privacy

- ... exploiting systemic economic excesses

- ... ... inflicting societal harm

- ... ... often with minimal upside to the consumer short-term

- ... ... or long-term

- ... colluding with large existing players

- ... ... to create a treadmill of declining product quality

- ... ... to facilitate government surveillance and information operations

- ... ... to "play defense" against competitors

It's kind of like how OOP was intended to stratify and contain developers, except instead of abstracting commodity PCs there's a whole different calculus which basically implies that the ultimate consumer of computing equipment is like, 2 or 3 big companies, because financing. You know how the old democratic saying goes, "none of us is as strong as the 2-3 permanent CEOs of reality."

The eye-opening number was 70% equity to deposit ratio. They were apparently betting the farm on the "there is no alternative" narrative.

I guess the huge leap of capacitive touch over previous technologies is kind of a footnote, but definitely bears mentioning here. Probably CEOs/COOs would have been plenty familiar with resistive touch which was much worse technology before 2007.

Gamecom, Palm, Casio would have employed this virtually unusable stuff, and the very idea of a touchscreen might have still seemed funny, if you hadn't have tried it yet.

Pressure is a differential phenomena.

You'll have to ask a smart guy what deposition, subtraction, injection, or teleportation process he's going to use to make a permanent diamond container that leverages the high compressive strength because I'm stupid.

Currently my chat history displays

> Not seeing what you expected here? Don't worry, your conversation data is preserved! Check back soon.

This is about 10,000x standard pressure. CRTs operate at 0.0000001x standard pressure, for reference -- they implode.

For example if superconductivity is desired from a 1 square inch (6.5e-4 m^2) footprint of this material, a weight of 6.5e5 N must fully rest on that footprint. Near the surface of earth that translates to a mass of approximately 146,000 lbs. Something like the mass of 3 tanks.

Did they dispose of their relationship with Crackle?

Looking, they sold at a gain in Dec 2020. Smart.

Apparently there's also SonyLIV in India, so not entirely forthcoming here. There's also Crunchyroll but that's niche, and Bilibili in China.

You're against transparent technology?

In the run-up to the pandemic as the yield curve inverted in late 2019 signaling broad expectation of economic trouble, SF Fed President Daly visits communities in the banking district.

See also Daly's report to New Zealand conference regarding inflation targeting, from the time period.

https://www.frbsf.org/economic-research/publications/economi...

In case you can't bring yourself to look at this, here's a highlight from the abstract

> [...] Our collective futures now include slower potential growth, lower long-term interest rates, and persistently weak inflation. [...]

Server software can run decades without intervention.

Imagine trying to predict the next two years of legislation that applies to you at the local, county, state, federal, and multilateral levels. It's hopeless because it's a system composed of the most complicated systems we know of -- people.

If you can't predict it, then controlling it is also hopeless. Control being one of the basic things a hacker seeks out.

Now if you convince a county of people to unanimously trust consumer electronics to process their vote fairly in a ballot measure, that and every facet about how you did that is obviously interesting.

Textual chads about what expressions of wealth inequality are egregious or harmful aren't terribly interesting or easy to read about.

I bring museums into the conversation because it is similar. Museums are often lavishly adorned palaces, largely serving the purpose of leisure.

To that it appears you may have fallen for, and are attempting to spread some discredited outrage porn/flamebait[0][1], possibly due to confirmation bias and the strength of your beliefs. Obviously since video was a fairly new media at this time, I'll agree that this attempt at depicting charity is a little more than off-color now, regardless of the precise contents.

I would check the HN Guidelines[2] since this whole comment kind of qualifies as fulminating and is kind of extremely political and ideological.

[0] https://www.politifact.com/factchecks/2022/sep/09/facebook-p...

[1] https://www.reuters.com/article/factcheck-queenelizabeth-foo...

[2] https://news.ycombinator.com/newsguidelines.html

It takes less resources to just say there's eternal suffering rather than actually manifesting it.

I'd love to hear your opinion on the perils of museums.

I mean unless you've got a better heuristic, but my conclusion is primarily structural with regard to Transformer units, basically that they're Turing-complete -- if there's anything more to reasoning and abstraction, I don't know myself, perhaps Scott Aaronson is the guy to talk to.

Right but the author's premise is that the AI fails to reason or extend ideas, not that the AI fails to reason or extend ideas in the presence of extraneous, conflicting, or malformed instructions.

If you're prompting ChatGPT (which is already loaded with prompt-engineering cheez-whiz mind you) to reason as a computer, why use words like "imagine"? why phrase things as a question? I think the fragility the author is observing here is extraneous to the actual underlying technology, and a factor of the UI that's been stacked-on and purpose-built for softer-purpose use, not hard academic research such as this.

Case study, even GPT2 can reason about the stack https://transformer.huggingface.co/share/bQXbmeyBZD

I would argue maybe that the author's premise depends on some faulty prompts.

With some gentle rephrasing, ChatGPT correctly interprets and manipulates the stack problem. I've observed ChatGPT may have some issues with non-standard sentence structure involving, especially with colons.

> Imagine a stack of items (from bottom to top) with a cat, laptop, television, and apple. The apple is moved between the laptop and the television. Which item is on top of the stack afterwards?

The author seems also to have misinterpreted that Stable Diffusion is the same sequence-sensitive model as Transformer LLMs, which it is not. While it is generative it has minimal capacity for abstraction, since the focus would rather be minimizing the "kernel" representation of each word in the weights.

DALL-E has a bit more luck, but requires much more verbose description to get anything like a screen on top of a cat, but that's presumably more due to the bias in the data set of things being displayed on screens, or cats getting on top of things.

Something like

> A cartoon image of a television above a cartoon image of a cat

https://imgur.com/q9NMY0W Illustrates that it's preference rather than capability.

Look, the developers went serverless, so now they don't have servers anymore they just have a monolithic website/app ecosystem, and any other access or attempted access is illegal systems intrusion, because they said so.

Never mind the massive Lazy Susan Scheme of unicast cloud bandwidth, collusion, and Berkshire Hathaway's money this whole activity entails.

Definitely the rule these days that any other driver doing something stupid is usually staring at their crotch, cupholder, or dashboard.

This seems like a legitimate middle-ground. Obviously quoting it would be -- kind of ballsy in an academic essay -- but drawing that line at least means that the cases where it's obviously awesome are (and should be) fair-game, even in formal writing:

- a thesaurus

- an organizational resource

- a creativity tool

- a summary expert that can lower F-K scores

Pretty effective at sneakily beaconing my location, presumably via its Bluetooth MAC through someone else's phone, and signing me up for an endless stream of advertisements for one specific pizza chain I live obnoxiously far away from.

Ah yes, using tmux so you can monitor your <endless search for for twin primes>.