guy98238710

Why? What's wrong with having others promote your ideas for free?

It's so sad and also terrifying when people call this real world programming.

Yes, soldier up, doggie! This world is not for pussies. Ditto for babies. Stop crying and man up! This world is for warriors. Fight and die!

> Even here in Germany, any blog, including private ones, need to come with an imprint section including your full name, residential address, mobile phone number et cetera.

How can this be consistent with GDPR? AFAIK, WHOIS data had to be concealed because of GDPR.

I would add a point to the pros:

- Even in the (very unlikely) case that nobody actually reads your post, language models will. They have to, because there's never enough training data. Your ideas will be merged into the text that LLMs synthesize and perhaps that will save someone's day.

> To release ideas that you have in your head and that you need to get out.

This. Getting ideas out of my head is the reason not only for blogging but also for all my opensource work. BTW, never sign an NDA. Locking ideas in your head will drive you nuts.

So perhaps a new UK-focused app that everyone sideloads?

One could easily argue that bullying in school is a much more widespread threat to children and the government is okay with that.

Isn't noise in the data going to dominate output size of lossless compression? Wouldn't linguistics and vision be better off with direct measurements of predictive strength?

This is such a great time to be a lawyer. Just play any AI like a slot machine until it gives you incorrect results and then sue. Probability is a bitch.

Marketing. OpenAI needs to create illusion that their chatbot is more than just gradual evolution of past models that still needs a lot of work.

Just pirate the books. You are only attracting attention of sharks when you pay.

Dogs don't get cavities. Human saliva is more acidic.

If you are very smart, chances are you will be able to obtain some power and influence. This is expected of AIs. Such effects can be however undone the same way you can undo effects of an explosion at a refinery (while punishing those responsible). The question is whether intelligence alone lets you expand power without bounds until you conquer the world. I just don't see how that would be possible.

> On your specific point, humans take decades to (very roughly) copy themselves. An AI could do it in minutes with far more precision and control.

Copying AIs just results in effectively higher intelligence (bigger brain). Things would be different with self-replicating robots of course, but duplicating robots is not that fast and one could argue that the robots are dangerous rather than the AI controlling them, because they would be dangerous even if controlled by stupid non-AI program.

> One day some other small hairless monkies want you to go extinct. Will they succeed.

The hairless monkeys were embodied and violent and they still took millions of years to dominate. The question is whether intelligence on its own provides sufficient advantage to be immediately dangerous. How can you conquer the world by doing "years of 200IQ thinking in an hour"?

Only if the body is duplicated too. If I could just make my brain bigger, I don't think my odds would improve much. Quite to the contrary. An AI with higher dependency on compute power seems more vulnerable.

I don't get this scare about AI. Imagine, you are a genius with IQ 200. You have studied hard your entire life and you are now an expert in several fields. One day you decide to conquer the world. Will you succeed or not?

There are some 100T to 1P synapses in the 20W human brain (of which we have 8bn). Every synapse is a little molecular computer with several kinds of memory and its own inference and learning logic.

AFAIK the biggest limitation of von Neumann computers is memory bandwidth. Brain is moving 10+ PB/s between neurons. This does not count synapse-local and intra-neuron bandwidth.

> Artificial Intelligence is moving fast

We wish. AI is hardware-limited and hardware is not moving fast. We are very, very far from matching raw compute power of the human brain. Robots are even more limited compared to human body.

Fine-tuning is like statically linking a large library (available in binary form only) to a small app. (An optimizing linker can blend app and library code so much they cannot be distinguished.) Weights intended for fine-tuning are still a binary, just a reusable one. Nothing new here.

Dunno who these guys are, but generally w.r.t. ML licensing, research kit is the important thing nobody is publishing. You know, the code used for diagnostics, visualization, benchmarks, decision support, etc. Research kit can be orders of magnitude larger than the final implementation code. Weights are just a binary with expensive build process.

The chart in the linked paper shows sharp drop between 20 and 30 years already. I remember another chart that I cannot find now, which showed wider age range and more components of intelligence and that one showed some peaks between 12 and 20.

Brain morphology is irrelevant. "Development" might just mean aging. Measurable performance is all that matters.

> Published work by experts in the field who have thought deeply about the subject, published works on the subject, and even designed and tested new curricula in a scientific manner.

This is called "paralysis by analysis", a special case of "penny wise, pound stupid". There's a lot of low-hanging fruit in education that does not need much thinking. You are free to optimize the solutions as much as you want, but don't use that as an excuse to delay change. Also, some changes (computers in particular) only bring benefits when implemented at scale. Small incremental tests will always show negative results, because they will be dominated by costs of the change.

> All I am asking for is for you to do even one single solitary ounce of effort in validating your assertions.

Homeschooling parents validate tons of possible changes all the time. I did it too. One thing you learn when homeschooling is that major changes can be implemented quickly, easily, and cheaply by single person. You can iterate so quickly you end up being limited by the compulsory curriculum rather than by opportunities for improvement. You then look at the glacially slow pace of change (or even regression) in "professional" education and cannot help thinking it's insane. Something must be very wrong with the system.

> But your whole approach of disparaging the very people trying to make a difference

Are they really trying to make a difference? My impression is that teachers just want to do the minimum work they have to do and then go home. I have never seen a teacher care the way parents care. There are exceptions, rare innovators, but exceptional teachers, if they avoid traps of educational fads and try to implement real change, quickly run into resistance. They are locked in the inertia of the large organization. Their hands are tied, they get frustrated and they eventually give up.

I am all for evidence-based decisions, but you are asking for evidence that sky is indeed blue. Like seriously, you want me to prove that math is actually used somewhere? How about it being used in the evidence you are asking for? And when was the last time you used long division? I am sure there are nuances that average student or parent cannot see, but the thing with schools is that they are so rotten that a long list of serious flaws is visible to the public along with corresponding solutions.

Your demands for evidence for the obvious are just a form of red tape intended to protect the status quo. This sort of behavior is the definition of bureaucracy: no action can be taken until approved by the bureaucrat (via curriculum, law, funding, or management decision). In the end, everyone in education just wants to be protected from the stress of change and the challenges of technology.

And BTW, it's not my job to show something is propaganda. It's curriculum designer's job to show everything in the curriculum is an undeniable scientific fact. That's easy to do with math or physics but much harder with humanities and arts, which is why all the propaganda concentrates there.

What religious fundamentalists? Maybe in the US? Where I live, schools are actually a source of aggressive religious propaganda. But the problems with schools are far wider than that they become ideological battlefields with propaganda-filled curricula. Lots of non-religious and moderate people see the problems.

Evidence of what? That doing arithmetic by hand like in 19th century (as opposed to arithmetic understanding/intuition/modeling) has zero applications in professional and personal life? Isn't that obvious? When was the last time you did long division by hand?

Or that mathematical modeling does have applications in both personal and professional life? Isn't it trivial to come up with numerous examples?

Or that mathematical modeling is so important it deserves place in the curriculum? Now that would be hard to show, but I never claimed it does. I just said it is relatively more valuable than pen-and-paper arithmetic, which is obvious from above two points.

Also, status quo does not get free pass. I imagine you would find it really difficult to defend pen-and-paper arithmetic, especially when everyone can see it's just bureaucratic inertia and technophobia.

This is for fearless kids. Most kids, I think, would be scared to death by these suggestions. It's an embarrassment minefield. It's better to start experimenting in a more private way in places where you are clearly welcome - student competitions, newbie-friendly websites, publishing stuff quietly without advertising it, reading and lurking, submitting tiny contributions that are clearly wanted, etc.

Kids these days do not suffer from scarcity of opportunities. They suffer from lack of confidence. They need to gain confidence through little successes. Sending them out into the world doing things way above their current skill level is going to result in confidence-crushing failures.

> Yes, don't we all hate it when people come into our job and tell us how to do it?

Teachers apparently need to be told, because they spend YEARS teaching kids how to do arithmetic on paper instead of handing out calculators and focusing on high-level mathematical modeling instead. There are lots of such insanities in the curriculum, not to mention loads of propaganda (not a US citizen, but AFAIK there is a lot of propaganda in US schools too).