bohm

there's no requirement to get your own domain name. when you sign up you automatically get a yourname.bsky.social domain.

Nope. You own your keys, and you can switch to a different server instantly while retaining both your username and your followers. They can't impersonate you because they don't have your keys. They can't ready your messages because they don't have your keys.

It's a bit more complex underneath (e.g. the are light-clients that you can entrust with part of your key), but that's the gist.

The posts and likes are all public. Some servers can charge users, and will likely have more resources to improve their offerings that way.

Nope, AT Protocol is not compatible with ActivityPub.

Yes, another reason to move away. Although Elon has also stated that he's tasked Twitter engineers with working on implementing E2E encrypted DMs.

I don't expect it to stay closed source. But there are significant differences to Mastodon. Just to give one example: Mastodon server admins actually can read your private messages. Also Blue Sky's approach to data portability and how to implement filtering/indexing is more developed. It might be personal preference, but I prefer AT Protocol's approach where the servers have less powers, and the users have more power. A lot of little design decisions add up to a qualitative difference.

OP here. I don't expect these servers to become unmoderated. I think a lot of them will moderate as strictly or stricter than Twitter currently is.

Author here. You get to have all the moderation, filtering, etc. that is implemented on the servers you use.

I'm sure there will be servers that very closely try to approximate what Twitter is/has been doing. Servers also can share any and all information needed to effectively moderate.

Do you have an actual critique of replacing centralized companies with decentralized protocols?

People still need transportation.

where's the problem for the consumer?

It's worth pointing out that YC startup foreverlabs.com does freeze your stem cells for precisely that reason.

Ok translation: Ivermectin, a safe and well-tolerated bacteria-derived insecticide already used in humans, can kill mosquitoes. By feeding humans ivermectin, mosquito populations can be reduced, which in turn reduces spread of the malaria parasite which is transported via mosquitoes.

take a daily herpes antiviral https://www.healthline.com/health-news/evidence-of-link-betw... and reduce your alzheimer's risk by 90%

We are off petroleum from Saudi Arabia. The US now is a net petrol exporter thanks to fracking. We have the cheapest oil in the world.

here's an alpha invite code: https://october.app/qr/TQSRG

We are actively developing https://october.app/ right now - check us out and give us feedback!

Gravity

self driving cars and uberbuses

guess it's time to disable flash for a few weeks...

From memory, but maybe a slight overstatement:

http://www.sci-news.com/genetics/article01036.html

70 per cent of protein-coding human genes are related to genes found in the zebrafish (Danio rerio), and 84 per cent of genes known to be associated with human disease have a zebrafish counterpart.

yeah agreed. i considered writing proto-prokaryotes.

Which is interesting considering we share about 85% of our genome with Zebra Fish.

According to the Santiago Theory of Cognition

"Living systems are cognitive systems, and living as a process is a process of cognition. This statement is valid for all organisms, with or without a nervous system." https://en.wikipedia.org/wiki/Santiago_Theory_of_Cognition

If this were a correct interpretation, even Bacteria could be seen as already incredibly intelligent. Going Zero to One (dead matter -> prokaryotes) is a significantly bigger step than going One to Five (prokaryotes -> humans).

This ignores that genetic code is meaningless without the translation machinery (ribosomes and the entire physical world of physics and chemistry). Genetic code without the translation machinery and the physical world it is translated into is meaningless.

A multi-cellular embodied intelligence can acquire information about the patterns of the physical world by growing in it - the code is compact because much of the necessary information can be acquired during embodied growth.

A disembodied intelligence such as AIs can't rely on external storage and acquisition of information during a cellular growth stage, and thus must likely be considerably more complex from the get-go.

Self-Driving cars will obsolete Univan Drivers

namecoin would fix this: https://www.namecoin.org/

I wrote about limitations on AI imposed by algorithmic complexity here: http://paulbohm.com/articles/artificial-intelligence-obstacl...

In essence: unless we extract that "single algorithm" that Andrew Ng believes in from an existing medium, we're unlikely to rediscover it independently.

(read the papers i link to at the bottom of the article for a more rigorous explanation)

I've worked on the Dropbox client back in 2009 and I can tell you that Dropbox has spent considerable resources optimizing Python, so there is actual python-related work to be done.

As Rian Hunter mentions they've for example implemented a custom memory allocator: http://highscalability.com/blog/2011/3/14/6-lessons-from-dro... and that's just one part of it.

With a few hundred million installed machines or so, having Guido on board to make python work better for Dropbox-scale apps is a huge win for them (and python)!