All of Vale's Comments + Replies

Vale10

A great collection of posts there. Plenty of useful stuff.

This prompted me to write down and keep track of my own usage:
https://vale.rocks/posts/ai-usage

Vale10

Predicting AGI/ASI timelines is highly speculative and unviable. Ultimately, there are too many unknowns and complex variables at play. Any timeline must deal with systems and consequences multiple steps out, where tiny initial errors compound dramatically. A range can be somewhat reasonable, a more specific figure less so, and accurately predicting the consequences of the final event when it comes to pass even further improbable. It is simply impractical to come up with an accurate timeline with the knowledge we currently have.

Despite this, timelines are ... (read more)

2Seth Herd
Please also consider the consequences of timelines WRT preparing for the arrival of AGI/ASI. From this perspective, accurate predictions of the technology and its consequences are very useful. For raw timelines, rring on the side of shorter seems much more useful in that if people think timelines are shorter they will for the most part be more prepared when it actually happens. Erring on the side of longer timelines has the potential to be disastrous. People seem to tend toward complacency anyway. Thinking they've got a long time to prepare seems to make it much likelier that we'll be collectively unprepared, and thus quite possibly all die. This isn't an argument to distort your timelines, just to make sure you're not overestimating, and to emphasize the real possibility of short timelines given the large uncertainty.
Vale30

Following news of Anthropic allowing Claude to decide to terminate conversations, I find myself thinking about when Microsoft did the same with the misaligned Sydney in Bing Chat.

1Amalthea
In the Sydney case, this was probably less Sydney ending the conversation and more the conversation being terminated in order to hide Sydney going off the rails.
Vale10

If many independent actors are working on AI capabilities, even if each team has decent safety intentions within their own project, is there a fundamental coordination problem that makes the overall landscape unsafe? A case where the sum of the whole is flawed, unsafe, and/or dangerous and thus doesn't equal collective safety?

Vale30

We have artificial intelligence trained on decades worth of stories about misaligned, maleficent artificial intelligence that attempts violent takeover and world domination.

Vale*0-1

I think people seem to downplay that when artificial intelligence companies release new models/features, they tend to do so with minimal guardrails.

I don't think it is hyperbole to suggest this is done for the PR boost gained by spurring online discussion, though it could also just be part of the churn and rush to appear on top where sound guardrails are not considered a necessity. Either way, models tend to become less controversial and more presentable over time.

Recently OpenAI released their GPT-4o image generation with rather relaxed guardrails (it bei... (read more)

Vale30

I went along to the VRChat meetup. Was absolutely wonderful to meet people, chat rational, discuss HPMoR, and generally nerd out for a while.

Thanks very much to everyone who organised events and helped with coordination!

Vale20

Speaking as a fellow Declan, I'm wondering if an unhealthy love for peanut butter is a "Declan-thing"...

Vale20

Really appreciate the pointers. I spend a lot of time lurking here but not much time posting. I'll definitely keep all your suggestions in mind next time I post. Cheers!

Vale67

if people had been playing with carbon fiber hair historically I really doubt horsehair would take off today.

I do oft wonder how many things we interact with are only the way they are because that's the way they've always been. Tangentially, how much of the world is just vestigial and nobody has thought to question it?

3ErioirE
Vestigial products and policies also tend to have an 'immune response', generated when parties who benefit[1] from the status quo actively resist attempts to change it. For example violin bow manufacturers could hypothetically fear lower sales if synthetic bows captured a greater market share, due to them not needing to be replaced as frequently. 1. ^ or even believe they benefit
Vale20

Personally, I've had my Caps Lock key bound to Escape for quite a while (I'd suggest a few years now?). It has been lovely and fits perfectly with my keyboard-driven workflow. Pairs great with Colemak-DH, which is my layout of choice.

I haven't found myself ever missing the Caps Lock key's original functionality, and my blank laptop keycaps mean I can have it bound without mental annoyance.

Vale10

The cost and time will gradually deter you and inherently create a scarcity mindset, sabotaging your creativity and playfulness and willingness to go "wouldn't it be fun if...?".

You caught something here with my own thinking. I have limitations associated with my own static site generator that acts as a simple barrier of just enough strength to deter my will and likely hinder output. I hadn't really concidered or thought of it at detail.

It isn't financial, but it is a time/mental tax I must pay to create, which should be eliminated. I'll aim to mentally emphasise it's credence as an issue requiring a solution.

Thanks for prompting the thought.