Nevermark

Not going to argue with the article on any front.

But when it comes to measuring productivity of transformational technology, a lot of things make that hard.

A lot of early use can be considered exploratory, whether it was intended to be or not. Meaning it was productive, but in a way that pays off down the road, not as much up front.

Not only are exploration payoffs not evident early, but they are also likely to be extremely high variance. Food for lots of "Well, of course it was a waste of time!" and "Well, of course it was going to pay off!" concurrent conclusions.

Then there is the fact that real productivity gains achieved early, may be the minority of wins, but are likely happening in places where they will grow very fast. At some point, they will "suddenly" appear and dwarf the losses.

I think the point was meant to be more simple and broad:

A system is not just its little bits.

Where "system" and "little bits" can be just about anything. But people endlessly repeat this mistake, and it has been a go-to misunderstanding/misrepresentation for a lot of people on this forum.

I did get that point! :)

(And you provided a valuable service/example for everyone who just thought the same thing! That phrase just keeps showing up. Whack! Whack! /h)

Unexpected free feature that plays right into recursive self-enhancement. What happens when they are trained to scale this ability up?

This could hit a wall, or be a step change in ... the rate of step changes.

The way I interpete it, is vast amounts of text are to LLMs, in some ways like vision is to us.

They can triangulate efficiently on the meaning from all the text at once.

It is why they are so good at processing ambiguity.

They can absorb an enormous amount of text and correlate the last thing that was said directly the first thing. That is much more like "seeing" than reading, or any other linear process - even though the field of vision is linear.

Next token prediction, if it requires solving math problems to get the next token is not "just" a next token predictor, its a machine that learned to understand math (as far as it was trained).

I am amazed people still don't know the difference between the shape of data (token prediction) and the actual problem (learning whatever is required to be able to predict the next token).

The former does not limit the sophistication of the latter in any way.

The dawn of relentless product/service bundling and then social media led the way to farming people, and capturing and gatekeeping people and businesses, instead of creating tools for people.

It turned out to be extremely lucrative.

Parasitical arrangements often are.

Because business law has not caught up with conflict-of-interest business models (surveillance-manipulation is that), and anti-trust law has not caught up with the unbelievable lock in that complex information-based products and services can leverage (network forces far greater than any form of price fixing, etc.), it still is.

The problem is worse than Silicon Valley losing its way. Civilization simply shouldn't tolerate parasitical businesses, no matter how many shiny yachts it's oligarchs let politicians vacation on.

Ethics are not a value to trade off against economic growth, they are one of the engines of both greater and less-centralized growth.

I am so glad I was already able to reconstruct true-to-the-originals C++ versions of Colossal Cave Adventure, Zork (mainframe version), Mystery Mansion and Warp.

Those are the ones that matter. :)

I don't see any company's blockade on this kind of activity having much impact. Too many models. Too much fun.

The appeal to a false dichotomy.

There is no reason that people who want to take on a ridiculous challenge shouldn't.

> ...you could build out sustainable settlements in the Antarctic, the depths of the Sahara...

That isn't an argument. I don't know what that is.

People who die on their way up to Mount Everest could .... ? People who spend their lives playing games could ... ? What a strange way to think.

Its good that different people really do want to do different things. Some of the best of have been taking on unbelievable challenges, pass or fail, live or die, since history began. Celebrate it.

I don't understand the projection of one person's reasonable lack of enthusiasm, onto the entire human race.

There are 8 billion of us. If a few want to make a go of it, you can be sure they will know the ridiculous challenges.

Space travel costs are falling rapidly. The Solar System has vast resources that seem impossible to reach today, but won't be for long in historical terms.

People still die climbing Mount Everest, for no purpose I can see. But I am not going to gripe if people continue making the climb.

I thought up a RAM that paired every data word, with an address word, pointing to its "next" word. ListRAM.

Reduces usable RAM in half, but speed is speed.

The idea being the RAM buses could be designed to provide fast pipelines of chained data to the cache, despite arbitrarily scattering in RAM.

There is no way this could compete with a single batch read of actually adjacent data. But it could speed up "Random" accessed data, which often is not - in fact - actually random from the program's or the data structure's point of view, just not adjacent.

And it would do one thing better: Reading a null terminated string wouldn't over read. The null next-address would stop every small string read from becoming a page read.

Even more flexible, send little read-chain algorithms to RAM. A little address math chaining and branching on in-RAM values, on little distributed address-math cores (RAM embedded APU cores!), could avoid a lot slow turnarounds.

These kinds of ideas seem even better to me now, by orders of magnitude, than they did originally. RAM really needs help.

> We either have free speech or we don't, there is no in between.

There is always a balance, always a line. Every priority exists in the "in between".

Nothing has 100.00..% priority.

Reality forces choices. Choices force legal, structural, cultural and (last, but unavoidable) contextual prioritization.

My guess is it is more like the opposite. They distribute spares evenly along every array in every dimension, so that there are always the same number of selected and unselected cores in each core-set.

With flexible addressing and bus-linking, along any set-line. So configuration happens the same way, with the same efficiency, regardless of whether any of the unselected elements are failed or not. As long as failures don't exceed the unselected-elements per element-array-set number.

And that number of unselected cores per element-array-set, is chosen to put wafer yields where they want them.

Trading a generous fixed wafer under-utilization for high yields, uniform configuration and processing efficiency, despite variable element failures.

Yeah.

An ambiguous threat of prosecution for corruption, hanging over the entire government workforce, with complaints accepted from anyone is a common tool for suppressing dissent and centralizing power.

As apposed to the practice of hiring professional auditors to systematically verify and strengthen compliance and transparency.

With any request, schedule a follow up meeting to discuss the completed work, or to perform an autopsy if there was some impediment.

Make sure to actually say “autopsy”.

“Inquisition”, “colonoscopy” or “employment review” all work well too.

If you have two Bob’s definitely get them involved.

Then have a practice of letting such meetings be cancelled if the work is done and went well.

This can be done as a consistent and transparent inside joke. Humor often generates the social permission to communicate hard truths, by also providing psychological stress dispertion.

I cannot vouch for any of this, but please share your experiences.

I would actually love a Sim City designed for realistic economic experimentation. A population of 1000’s of professions all acting self-interestedly with their various characteristics, motives, social and economic arrangements.

With a proper SimCity interface and the full complement of fun natural disasters of course!

Your point about past centralization is dead right.

But the increased ease of centralization in modern democracies that the Internet made possible is incontrovertible.

Which isn’t any special insight on my part.

Centralization has increased and decreased in many forms throughout history. There are many forces working in either direction.

I would also suggest that seeing more corruption in the open, is more likely to indicate more corruption out of sight as well, rather than the opposite.

What about insider trading by government officials?

Gifts from entities needing government approvals or under investigation?

Buying and selling of stocks in the same time frame as negotiation or pressure campaigns related to the same businesses?

Business deals with leaders of foreign countries arranged during diplomatic visits?

I am going to make $ millions when they release those hotlines!

The internet made coordination orders of magnitude easier than it could ever have been before.

Society and politics adapted to that, with like finding like, and reinforcing like.

So many effects, but one is local politics, which could be quite diverse across states, all converged into national politics.

Both a convergence of political views, and a convergence of cross-state, all-levels candidate and campaign coordination.

And the more coordinated a group, the greater incentives and opportunity for those at the center to further centralize and consolidate power.

The result can be a single person holding sway across all levels of politics for the nation, and a dilemma where most everyone in his party is either their because they are just as corrupt, or willing to hold their nose and concede some of their own power, as the price they pay to get anything done.

The Constitution has checks and balances within government to hold back centralization of power.

But no provisions to stop coordinated power blocks by unofficial organizations, even across all three branches, like parties can.

When coordination got cheap, fighting coordination becomes challenging and expensive. Centralization of power becomes a very powerful default.

> so what's next?

The automorphs. The quickly diverging types of self-designed, redesigned beings, that quickly start optimizing chemical processes, materials, biological cells, and everything at the superfast, supersmall scale on up, to maintain or enhance their survival expectations, in the presence of each other.

Redesigning cells from insanely inefficient (but still crazy fast) brownian motion networking, which allows competing objectives like building oneself vs. carrying out ones task to take place in the same place, to perfectly honed nano machines, pipes, conveyers, reaction chambers, composers, ....

Things already happen in the small scale many orders of magnitude faster than at our scale. Fine scale design of systems will add more magnitudes.

And then we come to it, ... the automorphs literally eat and re-eat their own substrates, and in the process eat the rest of the world. Materially.

> This anguish is especially common among on highly capable people, high achievers, and those used to some kind of ambition. And by accident of its history, there's a lot of those people in this industry and some of them are going nuts now.

I went nuts for years, until I sold off my interest in my dream partnership.

10 years of heaven. All the benefits and none of the downsides of my internal colleagues. Then 14 years of bureaucracy aimed at predictability over progress, which was infinitely more difficult for me to dodge as an external player.

Compensation for my interest was very objectively valued, with the rate of stalled progress priced in. Which made complete sense, while also being its own madness.

I walked away from so much value, but I don't regret it for a minute.

Humans are multi-objective beings.

Humans are built to focus on some objectives and not others at any given time.

Humans naturally prioritize objectives differently. Drives have exclusionary impacts on each other. In both real time and lifetime frames.

Reinforcing experiences further differentiate objective prioritization.

This leads people's behavior to be driven by entirely different net objective directions.

--

This diversity would be inevitable due to genetic diversity. But it is also a survival advantage.

Lots of studies show that populations maintain stable diversity of genes for entirely different life strategies, because operating differently from the majority is an advantage.

--

We don't actually have a "status drive".

We have drives that tend to generate status seeking behavior but can be subsumed by many different status-like pursuits, such as self-status of solving problems or completing single-person video games, being good at a hobby nobody else cares about, pleasing certain people even to the point of harming any recognizable definition of social status, acting to please (raise status with) internally generated or deceased personalities, a need to be contrarian or difficult to the point of low-social status but high internal status, etc.

--

Your assumption that same basic nervous system -> same net drives/behaviors, fails in much simpler species.

--

So no, mathematicians are no more motivation clones than they are genetic clones.

And mathematics is not a field that is attractive to people who give social status high priority, because it impedes social status.

Honestly, if you are trying to be funny you succeeded.

But if I am wrong, if you have the volumes of detailed inside knowledge on problems, sub-problems, and all their solutions, and alternate solutions considered, all applied to a car, but none of which are broadly or tangentially non-flying car related; just all the basic information that would be required to credibly make your claim, I will be just as delighted if you post the list.

I like Hover.com a lot.

Based on just my multi-year experience with a reliable interface.

So anyone else with a view on the company might chime in. But I am writing this because I have been very happy with them. None of the glitchy inscrutable experiences so many companies manage/love to design into their services.

Oh, the checkbox that turns off automatic renewal can take a second or two to register. So not perfect, but close. If that sounds trivial, you have understood why I like them.

This is well stated wisdom.

It also applied/showed half a century ago. Software is very developer dependent.

What that reveals isn't encouraging.

I think it is fair to say, the big order bits of software design quality - in general - have moved very little. Certainly less and less over the decades.

I do believe that where quality is very highly valued, there has been slow but valuable progress.

But everywhere else you look, software systems pervasively demonstrate obviously poor design.

That this may involve the unavoidable problems dealing with complex underlying systems, as much as it represents the application developer's lack of taste, isn't an excuse. It just demonstrates that poor design permeates the stack.

This is not an argument for perfection, but an observation that even slow mostly-monotonic progress has proven to be out of reach. So far.

There is a fundamental problem being unrecognized and unsolved.

These papers are just first steps. They are going to keep churning out papers, in part because the generated papers are going to get better. And better.

You are substituting one legitimate meaning of "worth" (moral value) with another legitimate meaning (economic value).

But then conflating the two to make your point, which if you care about making meaningful statements as apposed to word games, is not a legitimate move.

> Nobody's worth is tied to source code output.

If someone's expertise is worth nothing to those around them, they have a real practical problem.

In this context, the word "value" isn't referring to some moral model. It is referring to the asset value of skill.

Do you not relate to knowing whether a throw is going to swish or not, even as the ball is still just leaving your hand?

That's your analytical brain registering a signal from your intuition in a complex situation in which you have gained fluency.

Intuition drains many portfolios. But intuition that is informed by a lot of hard work and analysis has built many portfolios. Intuition doesn't always mean "no work", "guessing" or hopeful "manifesting".

> Yes, confident none of this belongs in a budget car.

Your confidence is empty.

Many "trivial" problems had to be solved to make this prototype. How many "smaller" ideas will end up in a budget car is something completely unknowable to you.

> a nerd — albeit a visionary one — who is socially awkward but got unfathomably rich and powerful

That part is true.

> manipulating democracy and being responsible for killing kids as a billionair makes his actions nefarious.

Indeed.

I think the admirable + despicable combination creates unprocessable dissonance for some people. People then take one side or another.