testmyname @testmyname@qoto.org

I love all of you and I want nothing but the best for each of you, particularly those on infosec.exchange. I understand that Mastodon isn't Twitter, that DMs are end-to-end encrypted, that we are spread across different instances and it can be hard to find your friends, and that an instance can go away at any time, and that translating posts doesn't work correctly, and there is no native giphy support, and that some instances are overwhelmed and super slow, and that you don't think the federated model can scale to a billion users, or that it doesn't support full text search of every post and account, or that we can't comply with the GDPR, or that we don't support quote tweet style functionality, or that we shouldn't collect IP addresses, and many other things.

The fediverse is a work in progress. I've been here for going on 6 years. In that time, it's come a long, long way. That said, Mastodon is not going to appeal to everyone. The decisions I make are not going to appeal to everyone. No one is forcing you to be here. No one is forcing you to disclose your personal secrets into a network of federated servers running by volunteers and hobbyists. NB: this is not Twitter. It has some similar functionality, but it is not Twitter. Parts of it are better, IMO, and parts are not. The security community is generally among the most skilled and competent IT people the world has to offer. Mastodon is open source. Do you see where I'm going?

I set this instance up a long time ago for reasons I don't even remember. I have poured my soul into this thing because I believe in the importance of this community. I have effectively peaked in my career as a CISO and I and my family live well. I am not running this instance for fame, money, a better job, or anything other than wanting to foster a community of people that can learn from each other and make the world a better place. That's it.

As I've said in several recent interviews, I felt particularly obligated to ensure the security community had a good handing spot in the fediverse as everyone was running for the doors in Twitter. We've grown from 180 active users to about 30000 in the span of 3 weeks. I do not expect everyone to stay. Some will set up their own instances. Some will move to one of the other excellent security focused instances. Some will give up and move to on to some other social media. And that is OK. While I am super excited to see the buzz here, I don't have subscriber targets, engagement targets, retention targets, or anything else. The only metric I hold myself to is whether I think this is serving a useful purpose to the community.

I appreciate all of you, regardless of where you land. Infosec.exchange has been here for a long time and will continue to be here for you.

As #mastodon have no algorithm for guessing for whom to deliver your posts, you have to keep in mind it when you write.

Meaning, use wisely these ingredients¹: target audience² (post visibility) and language, hashtags and content warnings.

[1] see image
[2] see https://nolog.link/s/69HCJt
#accessibility #tips #mastodonEtiquette

After two weeks in #mastodon my impression:
- it has a nice clunkly nerdy feeling. Focusing on the essentials, it feels like pioneering internet and I love it🔥
- The app is too much on the pioneering side😰
- my timeline is full of interesting people and interesting topic.💓 I'm not forced to skim through 50 tweets on the latest irrelevant news😒
- I'm getting more exposure. Maybe the AI doesn't filter me out as irrelevant, maybe people are more reactive, thanks 🙏

#introduction

Hello, Mastodon

The paper should hopefully be an easy read, without too much hard stuff and based on pedagogical examples.
It is a short introduction to a topic after all, my hope is to tantalize the brain and spark curiosity!

https://arxiv.org/abs/2211.06182

If you want to read more, look at the previous research paper 👇, here an explanation:
https://twitter.com/nuclearIdini/status/1339279400068665345

Let me know what you think!

(end)

Statistical learning studies 2)
If a model is too simple, it will not describe the system. If it's too complex, it will describe the data you have, but extrapolate badly.
When you build a model, you have to find a golden spot and that depends on how many data one has & expect

Only very few nuclei exist on earth and in the lab. With these few we want to constrain complicated models to incredible levels of precision. This requirement is not always reasonable.

(3/n)

The model should:
1) do few enough operations for the calculator to output a number in time.
2) have little enough complexity so the data can constrain it.

Computational complexity studies number 1).
Some models are easy, others hard i.e. require a lot of operations. Quantum mechanics is hard. It is hard even for quantum computers.
Consider this when building a model!

(2/n)

New proceeding paper today on arXiv at the intersection between the theory of computing, machine learning, and nuclear physics.

An introduction to computational complexity and statistical learning theory applied to nuclear models
https://arxiv.org/abs/2211.06182

a short tootorial🧵

Basically: how do we model things in physics? We need calculators & data to tune our model to the real world.

(1/n)

Mastodon: an introduction for beginners and for scientists.

https://giorgio.gilest.ro/2022/11/05/mastodon-an-introduction-for-beginners-and-for-scientists/

Ukrainian people are holding up the torch of freedom for all of us.

Ukraine belongs in the European family.

#StandWithUkraine 🇪🇺 🇺🇦

Back at it again with the free stuff~

I’m giving away a 1-year IVPN (ivpn.net) Pro subscription to one of my lovely followers who boosts & replies to this post. A winner will be randomly selected in a week, good luck!

IVPN is one of the providers we recommend at @privacyguides. I’m/we’re not sponsored/affiliated with them in any way, I just like their service 😁

Seeing a lot of spreadsheets going around with people to follow on #Mastodon, which is super cool, but y'all should check out and add yourselves to #Trunk, a site with a bunch of different lists which has been around on the fediverse basically forever: https://communitywiki.org/trunk 😄

#TwitterMigration #list #follow