Andrey Sozykin

[Andrey Sozykin @asozykin]

@apexlearn_org @LucaColonnello @kiknaio @jenzhuscott It's not clear what you mean by the "paradigm shift." Where is the shift between before and after the web?

Ongoing search for an optimal information exchange rate between regions, including filters, etc, but life goes on as usual, governments just do their routine, right?

[Andrey Sozykin @asozykin]

@danilobzdok @KordingLab @bryan_johnson @theallinpod @friedberg Substance can't provide experience.

Substance doesn't carry information. What you call "experience" is just sparks from a broken brain.

Substance is not a book or a film. It is just noise that deletes experience. It can only be a medicine for an already broken brain.

[Andrey Sozykin @asozykin]

@ViralVideos @elonmusk It's not whether you believe in free market or not.

Free market creates new things, government planning maintains old things.

These 2 are inseparable like Yin and Yang.

The key to success lies in the optimal distribution of resources between market and gov planning.

[Andrey Sozykin @asozykin]

@NeuraNova9 @grok The one on the right is better, but the left one is more on the topic

[Andrey Sozykin @asozykin]

@fchollet Even if software updates were free, some would still have negative value, e.g. design changes for an already well-designed GUI.

[Andrey Sozykin @asozykin]

@fchollet Are you sure there's still latent demand for software?

New software has to replace older software because the user's attention span is limited.

Then, the users have to learn to use the new features.

Does the value of new features outweigh the cost of learning?

[Andrey Sozykin @asozykin]

@elonmusk We need >1 separate loosely connected civilizations in order to exceed the expected lifetime of one. The more isolated the better. Similar to cell division, in a sense.

We'd be doomed if information could've spread without limits, but there's the speed of light.

[Andrey Sozykin @asozykin]

@mioana @KordingLab Handling of the entropy of supply and demand is the key to prosperity: what is being produced/consumed is important, not how much. What we pay our limited attention to. But this would be a different model. If I ever get to finish it, will definitely make a similar demo)

[Andrey Sozykin @asozykin]

@mioana @KordingLab Wow, thanks for the interesting model! These live notebooks illustrate ideas better than any presentation. Neuromatch-style.

I wonder why anyone would care about capital and labor while getting extinct, though. But this doesn't take away from the greatness of the design :)

[Andrey Sozykin @asozykin]

@fchollet You have a weird image of Russia. A man is shot every hour in the US, 40 million are in poverty, and a comparable number addicted to drugs. Russia is actually a paradise compared to that. It's a highly developed european country.

[Andrey Sozykin @asozykin]

@fchollet It's not from scrolling I'm afraid

[Andrey Sozykin @asozykin]

@ChrSzegedy @fchollet For the most general abstract "idea" the value is 0 (no free lunch theorem).

For a set of ideas conditioned by human priors - positive expected value (e.g. science).

For AI generated code - we'll see, but can very well be negative.

[Andrey Sozykin @asozykin]

@ChrSzegedy @fchollet Not sure if the mental image itself is an asset.

This image is a model, and only a subset of models are useful.

An idea requires a real world implementation. Implementation has a guaranteed cost + a chance of returns. Expected value of a generalized idea can be net negative.

[Andrey Sozykin @asozykin]

@elonmusk In order to run a project on the moon, one needs approval from the big nations. They can ruin the whole thing otherwise. How do you plan to come to an agreement with competitors given the current state of the world?

[Andrey Sozykin @asozykin]

@KordingLab @mioana There's also an explanation for why the economy gets worse and worse despite innovation, but that's another story of course

[Andrey Sozykin @asozykin]

@paulg There should be a natural repelling force pushing the independent-minded apart.

The sources of ideas have to be distinct enough for the conventional-minded to follow and eventually test these ideas.

Clustering together is not the way for the independent-minded, loneliness is.

[Andrey Sozykin @asozykin]

@fchollet It can actually be worse than 10% as USDX only compares USD to 6 "major" currencies, most of which in turn heavily depend on USD.

Less USD-dependent BRICS currencies like RUB, INR and CNY would constitute a better USD index, in my opinion.

[Andrey Sozykin @asozykin]

@elonmusk @fchollet No such deadline is possible.

Every new tech has liability (maintenance, excessive function) alongside the benefits it brings. Net value is netto negative for many projects that get implemented anyway.

Higher tech means more control of the project efficiency, more regulation.

[Andrey Sozykin @asozykin]

@davidpattersonx @Brueck1988 @elonmusk @fchollet Robots allow for higher quality and scale of production. But in most cases, these are not limiting factors - there's no demand for higher quality/volume. In such cases, robotization will lead to loss of production efficiency and higher prices as it already happens with software.

[Andrey Sozykin @asozykin]

@davidpattersonx @Brueck1988 @elonmusk @fchollet Most modern software projects make businesses less efficient due to inadequate measurement of project efficiency:
1. Managers initiate unnecessary projects for their career needs
2. Software providers sell excessive functionality
3. Updates are new code without new fictionality