Front-end developer, designer, writer, and avid user of the superpowered information superhighway.
Vale
Karma: 88
I appreciate the assumption of preconceptions, but I began having a poor experience and then discovered the post discussing the move. It was while trying to figure out why things were acting up that I found the announcement of migration.
When I first discovered LW 2.0, I was blinded by its majesty. It was beautiful. However, I can’t help but feel some of that beauty has been lost.
I feel even more has been lost in a move to Next.js and the horrors of Vercel. I regularly find myself staring down the barrel of a black void with the all too common text: ‘Application error: a client-side exception has occurred (see the browser console for more information).’
So many little bugs and glitches that never reoccur twice but are everywhere. This isn’t even something I can report, because the issues are everywhere and appear inconsistently. Most of the time they can’t be replicated, but I figure most people using this site know what I’m talking about. The framework special of clicking a button and nothing happening, then clicking it again and the initial action happening, and then the second input suddenly taking place immediately after, by which point you’ve already clicked the button again and now things are in some half-broken state until you click away and then back.
I’m certainly not alone in these thoughts.
I understand that these actions have been taken in an effort to make the site easier to develop and more in line with the tech used by the rest of the projects the team works on, but the user experience is abysmal. I don’t personally love the experience or appearance of GreaterWrong (due to personal preference, not any particular issue with the site), but it seems like the only way to read LessWrong without experiencing constant breakages and interruptions.
LessWrong was such a beautiful and well-crafted website, but it seems to only stray further and further from that polish. Every action taken seems to push the site further from the web platform and further towards using some reinvention of the wheel. Please reconsider the choice of technologies and whether the recent changes are worth it.
To me, this sounds like you’re simply pushing the problem a little bit downstream without actually addressing it. You’re still not verifying the facts; you’re just getting another system with similar flaws to the first (you). You aren’t actually fact checking at any point.
I just saw the term ‘Synthetic Intelligence’ thrown forward, which I quite like.
Many people agree that ‘artificial intelligence’ is a poor term that is vague and has existing connotations. People use it to refer to a whole range of different technologies.
However, I struggle to come up with any better terminology. If not ‘artificial intelligence’, what term would be ideal for describing the capabilities of multi-modal tools like Claude, Gemini, and ChatGPT?
We talk and think a lot about echo chambers with social media. People view what they’re aligned with, which snowballs as algorithms feed them more content of that type, which pushes their views to the extreme.
I wonder how tailor-made AI-generated content will feed into that. It’s my thinking and worry that AI systems can produce content perfectly aligned with a user in all ways, creating a flawless self-feeding ideological silo.
I was thinking a little bit about the bystander effect in the context of AI safety, alignment, and regulation.
With many independent actors working on and around AI – each operating with safety intentions regarding their own project – is there worrying potential for a collective bystander effect to emerge? Each regulatory body might assume that AI companies, or other regulatory bodies, or the wider AI safety community are sufficiently addressing the overall problems and ensuring collective safety.
This could lead to a situation where no single entity feels the full weight of responsibility for the holistic safety of the global AI ecosystem, resulting in an overall landscape that is flawed, unsafe, and/or dangerous.
Taking time away from something and then returning to it later often reveals flaws otherwise unseen. I’ve been thinking about how to gain the same benefit without needing to take time away.
Changing perspective is the obvious approach.
In art and design, flipping a canvas often forces a reevaluation and reveals much that the eye has grown blind to. Inverting colours, switching to greyscale, obscuring, etc, can have a similar effect.
When writing, speaking written words aloud often helps in identifying flaws.
Similarly, explaining why you’ve done something – à la rubber duck debugging – can weed out things that don’t make sense.
I don’t necessarily believe or disbelieve in the final 1% taking the longest in this case – there are too many variables to make a confident prediction. However, it does tend to be a common occurrence.
It could very well be that the 1% before the final 1% takes the longest. Based on the past few years, progress in the AI space has been made fairly steadily, so it could also be that it continues at just this pace until that last 1% is hit, and then exponential takeoff occurs.
You could also have a takeoff event that carries from now till 99%, which is then followed by the final 1% taking a long period.
A typical exponential takeoff is, of course, very possible as well.
Extremely quickly thrown together concept.
There is a tendency for the last 1% to take the longest time.
I wonder if that long last 1% will be before AGI, or ASI, or both.
OpenAI’s Jig May Be Up
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
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 popular – both with the general AI hype crowd and those more informed. People don’t seem to penalise incorrect timelines – as evidenced by the many predicted dates we’ve seen pass without event. Thus, there’s little downside to proposing a timeline, even an outrageous one. If it’s wrong, it’s largely forgotten. If it’s right, you’re lauded a prophet. The nebulous definitions of “AGI” and “ASI” also offer an out. One can always argue the achieved system doesn’t meet their specific definition or point to the AI Effect.
I suppose @gwern’s fantastic work on The Scaling Hypothesis is evidence of how an accurate prediction can significantly boost credibility and personal notoriety. Proposing timelines gets attention. Anyone noteworthy with a timeline becomes the centre of discussion, especially if their proposal is on the extremes of the spectrum.
The incentives for making timeline predictions seem heavily weighted towards upside, regardless of the actual predictive power or accuracy. Plenty to gain; not much to lose.
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.
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?
We have artificial intelligence trained on decades worth of stories about misaligned, maleficent artificial intelligence that attempts violent takeover and world domination.
Vale’s Shortform
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 being able to generate political content and images of celebrities without consent). This came hot off the heels of Google’s latest Imagen model, so there was reason to rush to market and ‘one-up’ Google.
Obviously much of AI risk is centred around swift progress and companies prioritising that progress over safety, but minimising safety specifically for the sake of public perception and marketing strikes me as something we are moving closer towards.
This triggers two main thoughts for me:How far are companies willing to relax their guardrails to beat competitors to market?
Where is ‘the line’ between a model with relaxed enough guardrails to spur public discussion but not relaxed enough to cause significant damages to the company’s perception and wider societal risk?
When reflecting on the past, I, like many others, cringe. However, I’ve come to consider this not as a source of regret but as a positive signal of growth.
I once heard the perspective that cringing about the past indicates growth from that time. You’re identifying that there are things you did at that point which were regrettable and which you would endeavour to avoid now. It is representative of the difference between your current self (your updated models, values, and social calibration) and the past self who performed the offending action.
Much of the time, things we look back and cringe about now we did not find cringe-worthy at the time, indicating a change has occurred.
Thus, cringe works somewhat as a measurement of growth. If you do not cringe at all looking back at past actions, then it implies one of two things:
Your past self was remarkably optimal and well-calibrated.
You haven’t significantly updated your models/values or changed since then, and you are unable to identify your past flaws.
I consider cringing as valuable data evidencing that self-correction and learning mechanisms are functioning.