reroute22

It was emotional for me to watch this. I think it's tragic, what's happening. I'm not a professional mathematician, but I have a degree in a program where the first 2 years were practically the same with pure math program, I totally get it. I have professional mathematician friends.

I haven't spent all my 20s on a single proof, but I have spent all my 20s on education, professional immigration, and building a certain chip for a certain well known computer company. I have lost and sacrificed important things along the way. It's an emotionally charged topic for me.

I feel for the people affected.

Stay strong, you are not alone.

I agree in both parts. In particular, repeatability is only one part of it, another is in person physical presence, or a weaker form of it - on-screen presence, or at the absolute least - voice.

All of the latter plays a huge role, a lot of us don't want to be horrible to a person physically in front of us, evolutionally it makes sense - if you do, you might get a fist in the face. Or no support and help in the future, as whoever is nearby will have a high chance of meeting you again (historically, before modern technology and transportation), or even belong to your tribe.

Notice, however, that all of these factors in the end correlate with pre-existing geographical proximity to the person in question. In fact, the said proximity also maximizes the likelihood that you'll see eye to eye - will not have radical, difficult to overcome differences in world view. Yes, sure, we all "should" understand people existing in very different circumstances, but let's face it, it goes a lot better when circumstances are similar, and that once again correlates with geographical proximity.

So realistically there is a huge correlation between "connecting person A and B over the internet is good" and "person A and B already live nearby and already have a high chance of running into each other physically in the real world".

In its simplest, most primitive form what you want is to give people who have already physically met once a way of continuing the connection even if circumstances drive them apart immediately after. And if possible, keep that connection to more physical-like forms - at least audio, better yet video calls, only if absolutely necessary - text.

Wait, that's called a phone. Invented a gazillion years ago. You meet first, then you exchange the numbers. And then you call and actually talk to the person.

What social networks are doing is a fundamentally problematic way of connecting people even without any recommendation systems, rage baits, and other social and addictive engineering involved, or ads - even in the purest form it's already a problem.

Execs was indeed more of a metaphorical description.

A small subset of individuals working there.

You don't need 80k people to make policy decisions, Meta is that big not because of policy makers, but because of infrastructure people and people working on all sorts of Meta's attempts to enter other markets to expand their business, markets that have nothing to do with addictive and/or exploitative social networking products.

A president of a country also can't do anything by themselves, they need a small army of supporters in positions of influence. And still the said army is a percent of the total population.

Besides, let's be real here, yes, Meta's is one of the worst in terms of overall impact, e.g. certainly worse than Google out of the big ones, but the difference is not as big as one might imagine.

Meta also powers WhatsApp, which is basically an operating system of the entire region of Southeast Asia and India, and also a large portion of South America, which together host >2.5B people. Loosing that particular part of Meta (if Meta was to fail) would be a big loss to humanity, loss that Western dwellers refuse to know or accept, but ask anyone living in those areas they'll tell you how much of the economic boom of late in those places rides on connectivity provided by WhatsApp.

Google on the other hand owns YouTube, and in my personal anecdotal experience I get a heck of a lot more misery from YouTube and especially Shorts that I can't seem to escape or avoid, than I do from TikTok or Instagram neither of which I use at all and have no issue ensuring 0 interactions with.

Yes, in the end both do both good and bad, and yet in grand total Meta is worse than Google, but A) the difference isn't as large, and B) it's actually a lot harder to tell than it seems if the net result is even negative in the end. It probably is, but it's hard to tell definitively. As in, stating that some parts of Meta's business are better die and the humanity will be better off is undeniably correct, stating that all of it just needs to seize to exist sounds more like a hyperbolical illustration of sentiment, not an objectively sound proposition necessarily.

While reality can be anything but.

As far as I understand, there is plenty of research there in disciplines raging from social studies through psychology to game theory and economics, as well as informal simulations, that strongly suggest that human interactions are positive to participants pretty much if and only if those interactions are repeated, which realistically only occurs if participants are circumstantially close already - same neighborhood, same job, family, friends, same school, etc.

One-off interactions are almost invariably toxic with at least one of the participants getting cheated, bullied, or otherwise harmed.

So the whole premise of connecting people unconditionally, including anonymously, automatically, and from opposite sides of the world is inherently broken and doomed to do a lot of damage.

So even Meta's self proclaimed mission is damaging to society if followed, what could possibly at that point be expected from what they actually do, given the combination of basic facts that the primary purpose of any business is to make money, Meta's specific notoriously evident disregard towards ethics, their position as an advertisement business and entertainment provider, being deep into enshitification and market saturation, and of course actual honest mistakes to boot.

Don't confuse employees with execs. It's a gigantic company with almost 80k employees.

Most cultures around the world are acutely aware that the actions and opinions of their leaders are not a reflection of behaviors and opinions of regular citizen.

Such a fantastic paper overall, it was a pleasure to read, it's very accessible, and greatly informative. If anyone is new to the idea and is seeking a definition of AGI, reading this paper is easy and is immeasurably superior to merely googling or reading the wikipedia article.

My only criticism for the article within the particular set of goals outlined above is this:

The paper seems to be under-exploring two aspects that appear to be worth exploring explicitly and in detail:

1. Ability to rapidly learn from a very limited amount of instructory data post-deployment and substantially advance in its abilities in the domain of the learning post-deployment, as opposed to possessing certain level of professional skills immediately on-deployment.

2. Ability to invent entirely new ideas, like for instance inventing an entirely new system of numbers or another symbolic or other system on its own, all to advance its current goals.

Both in part to distinguish an AGI from a large collection of glued together Narrow AIs, each purpose-built for a specific, but entire domain of fairly loosely related tasks, and in part to ensure that a high level AGI system always appears at least as intelligent as an average human teenager across the full spectrum of all possible cognitive and metacognitive interactions with the said teenager (be those interactions initiated by another human or by the cognitive projection of the environment).

Without these abilities, there could be a system - it could be argued, I believe - that would technically (or at least arguably) satisfy the definition of an ASI level of AGI as per the paper that an average human child / teenager may appear more intelligent in comparison to, exceeding the said system in plasticity and real-time limited-input adaptability of the intellect rather than off-the-shelf proficiency in trained adult human tasks: a high level AGI system might be initially trained on trillions of tokens of input data, but once deployed, it needs to be able to acquire new skills and proficiencies from mere tens-to-thousands of input examples - such that humans do.

Perhaps the framework presented by the paper intended to silently encompass these abilities and the remarks above, but surely they deserved a separate discussion, such as other aspects of the definitions and the framework proposed by the paper are indeed explicitly discussed.

Similarly, not including "autonomy" into the "six principles" (making them seven) for composing a definition of an AGI and only discussing it briefly and on a side also appears to be a questionable choice for the same reasons.

I'm afraid I'm with your colleagues.

I value being able to grasp what the code does at a glance while scrolling through. Your first snippet is entering my brain like as if without even reading at all, all at once and instantly.

Your second example absolutely does not and I'm sure you know why.