Não vejo a hora de saírem os primeiros estudos sobre o impacto no uso de IA a longo prazo, principalmente no trabalho, em pessoas com TDAH. Acredito que vai piorar sintomas como distração, ansiedade e estresse, e eventualmente burnout. :( Inclusive em neuro típicos!

I have been off social media for most of the previous month due to personal issues. Large breaks and sparingly used. Feels good. Now I need to turn this into a habit, lol. :-)

¯\_(ツ)_/¯ github.com/microsoft/ty...

I hope Brazil will make history tonight.

If you don't trust your employees to work remotely then you probably hired the wrong people and that is on you.

Hey, I created a Bluesky feed for database hackers and related folks. Please let me know if you would like to be included (or removed) from this feed. :) bsky.app/profile/did:...

Some days feel longer than others. Some days are busier than others. Some days are tougher than others. Today was one of those busier, longer, tougher days.

Jerry Lawson invented the video game cartridge. He's the reason they stopped making pong consoles in the 1970s. If it weren't for him, the Atari 2600 and many others would literally NEVER exist at all.

After DeepSeek, markets are realizing that LLM training can be done more efficiently. But I think new markets will appear for private LLMs and local inference. Chips will still be needed but with more dispersed buyers. (I don’t know anything about markets, so time will judge this guess)

Quão vira latas as pessoas tem que ser pra pegar um vôo longo e ver a posse pela TV?

The great news is that you don't have to wait for Santa 🎅 to bring you a physical copy this holiday season. The book is available right NOW for download from Dremio. Check it out! https://hello.dremio.com/wp-apache-iceberg-the-definitive-guide-reg.html

Vector database companies will become one of the three a. Search platforms (e.g., Elastic) b. AI platform (inference/training/pipeline) c. OLTP provider (e.g., MongoDB, Postgres) Hard for vector databases to be a separate offering/company

I gave a talk today in which I walked through key points from our Bluesky architecture paper, and got some good questions www.cl.cam.ac.uk/research/sec...

Reconnecting with the Twitter homies on Bluesky

Hey, @jay.bsky.team's interview on CNN yesterday night: edition.cnn.com/2024/11/20/b...

Buying a book and actually reading it are two very different hobbies.

One way we know Bluesky has the juice is because the starter packs used to be super generic like “Environmentalists 😀” and now they are super specific like “North Atlantic salmonid fisheries scientists whose names start with T-Z” and they have the full 150 people in them

@dep4b.bsky.social Hey, man. Nice to see you here! ☺️ It's very cool to see IR/search folks getting together at Bluesky.

I blogged: http://blog.jnbrymn.com/2023/05/20/primary-objective-of-software-design.html The Primary Objective of Software Design: Minimizing Total Cognitive Load

Hey @github.com users on @bsky.app, did you know you can get a github.io handle in a few steps? 1/3

Sunday mood.

It is always Wednesday. :(

Very good timing for the Builders of AT Protocol tech talk this Thursday! lu.ma/lhuj8gyh

R.I.P. Quincy Jones, um gigante entre gigantes. (1933-2024)

NGC6302 Butterfly Nebula

2024-10-28

New post! We open sourced SlateDB a week ago. I wrote down some notes about its origin, what it's good for, and where it's headed. And, of course, the obligatory Github ⭐️'s vanity metric. 😃 materializedview.io/p/slatedb-an...

Indeed.

Feynman plays the Bongos youtu.be/qWabhnt91Uc

Bom dia!

Concordo. Inclusive tem aquela métrica que mede essa profusão de if-else aninhados: a complexidade ciclomática. Softwares de análise estática de código como SonarQube pegam isso. 🙌🏾🙏🏾 en.m.wikipedia.org/wiki/Cycloma...

😞

You were born with a super computer the tech industry is spending trillions trying to replicate. Train your own model. Put that real intelligence to work.

O divo tem até música m.youtube.com/watch?v=4L0D...

Por sinal, o bottom up DP não tem nem recursão nem Memoization, usando tabulação dos resultados. ;) O exemplo clássico de bottom up DP é o Fibonacci onde se popula os resultados num array começando de arr[0] até chegar ao N desejado. Ou usar duas variáveis.

Exatamente o que Matheus bem descreveu. DP vai além, pois bottom up DP, por exemplo, nem usa memoization nem recursão, mas tabula os resultados. E pra além disso, os problemas de DP devem ter sub-estrutura ótima e sobreposição de subproblemas.

Não. Cache é uma técnica. Utilizada na DP e em Backtracking com o nome de memoization. Mas que é utilizada em outros contextos/problemas. Pra classificar como DP/Backtracking o problema tem que ter natureza recursiva. É sobre recursividade.

Programação dinâmica é mais do que só "guardar e acessar rapidamente" com cache. Embora o uso de cache seja parte disso. O lance está em identificar subproblemas, definir relações de recorrência e estruturar a solução de forma + eficiente.

Bluesky now has over 10 million users, and I was #5,430,884!

He’s so close to realizing that everyone in his life only pretends he’s funny to keep the money flowing

Jura?

Todo fã de bilionários como Musk e Bezos é, no fundo, um desesperado. O sujeito defende a existência dessas aberrações do capitalismo em nome da vã esperança de que a sua própria miséria seja redimida, de que um dia ele também se torne um bilionário. Mas no fundo ele sabe que nunca será.