AI

[kipo]

AI ce zamijeniti mene, tebe, ribu influenserku koja tek pocinje uzivati prilike kojima je odusevljena. Kroz 10 god bice strasno. suzivot i iskoriscavanje AI. Kroz 20 god bice dobro, ako ipak bude bioloskih jedinki i ne budemo matrix. Kroz 30 god ili vise - vjerovtno ne - tek ako Maskove neurolink elektrode prodju kao jednini nacin da ispratimo brzinu AI. Posle toga, sve je do AGI.

[Yoyogi]

4 hours ago, kipo said:

AI ce zamijeniti mene, tebe, ribu influenserku koja tek pocinje uzivati prilike kojima je odusevljena. Kroz 10 god bice strasno. suzivot i iskoriscavanje AI. Kroz 20 god bice dobro, ako ipak bude bioloskih jedinki i ne budemo matrix. Kroz 30 god ili vise - vjerovtno ne - tek ako Maskove neurolink elektrode prodju kao jednini nacin da ispratimo brzinu AI. Posle toga, sve je do AGI.

Ima to drugi potencijal. Neko ko te mrzi, moze da kupi drona za 50EUR, stavi mu tvoju sliku, da mu okvirno adresu gde zivis i posalje ga da te ubije. Dron se samounisti posle toga. Ti mrtav i nikome nista.

[Downforce]

I sve sto treba je da kaze dronu da pronese velikom brzinom poruku kroz tebe...

 

U slucaju da dron kaze da je nono ubiti coveka.

 

 

[eumeswil]

On 18. 6. 2024. at 11:49, Caligula said:

 

 

Tako je - iskreno me iznenadila navedena cinjenica (?) da na procesore AI odlazi 2% proizvodnje svetske elektricne energije i da se za hladjenje koriste trilioni litara vode, od toga trecina pitke vode... dalje, neisplativost: Midjourney samo na mom spratu u nekom habu u BG/Srbija koristi nas cetvoro sa pretplatama od 30-60E. Milion dolara mesecno za odrzavanje MJ je smesna cifra, a za autorku je katastroficna.

 

I tako dalje - o moralnim i drugim zakljuccima da i ne zapocinjem, ovo njeno mi lici na senzacionalizam.

 

 

Jeste, taćno je za vodu i za energiju generalno.  Pogledaj ovaj dva videa.

 

 

 

 

 

Google, Meta i Microsoft su zajedno u 2022. povlačili 2,2 milijarde kubnih metara vode što je dvostruka godišnja potrošnja vode države Danske.

[Yoyogi]

2 hours ago, Downforce said:

I sve sto treba je da kaze dronu da pronese velikom brzinom poruku kroz tebe...

 

U slucaju da dron kaze da je nono ubiti coveka.

 

 

To je vlast koristila, face recognition, na onim protestima u Sapcu. Ljudi su, stotine njih, sutradan dobijali prijave da plate kaznu a da ih niko nije ni zaustavljao, trazio licne karte ili ih ista pitao.
Das sliku dronu i napuckas ga, to je i sada tehnicki moguce samo su dronovi jos skupi. To mozda za velike zverke, kao sto su Izraelci ubili onog vodju Hamasa u sred Teherana.

[bags]

Danas ucestvovao sa svojim timom na AI hackatonu.

 

Ja sam otisao prilicno sa niskim ocekivanjima ali sam ostao pozitivno iznenadjen.

Svui timovi su dobili generalnu temu i onda je svako mogao da ubaci svoje elemente. Mi smo recimo imali ideju da kad dodje novi developer u tim odgovori na par pitanja o svojim prethodnim iskustvima i onda dobije iz Jire spisak otvorenih bugova sortiran prema vise kriterijuma. Koristili smo za input tekst i komentare ticketa i naravno vec zatvorenih bugova. Kljuc  je bio sto smo uspeli recimo preko PR-ova i komentara razumeti o kom delu stacka se radi i onda preporucimo. Znaci ako si izabrao da si radio sa frontendom mi cemo proci kroz otvorene bugove, potraziti slicne zatvorene bugove (koji ukljucuju rad sa FE) i onda odabrati slicne i odraditi summary.

 

Vazno je istaknuti da je za uspeh najvaznije bila priprema organizatora. ( pripremljen template za Python projekat i config za VS Code sa continue.dev da ti olaksa programiranje)

Druga bitna stvar je bila mali timovi od 3 clana gde svi mogu programirati.

 

Prosli smo neocekivano u sledecu rundu medju 10 najboljih i sad se moramo bolje pripremiti za sledecu rundu i dodatno pojacati AI.

[Lord Protector]

Quote

 

Learning to Reason with LLMs
We are introducing OpenAI o1, a new large language model trained with reinforcement learning to perform complex reasoning. o1 thinks before it answers—it can produce a long internal chain of thought before responding to the user.

Contributions
OpenAI o1 ranks in the 89th percentile on competitive programming questions (Codeforces), places among the top 500 students in the US in a qualifier for the USA Math Olympiad (AIME), and exceeds human PhD-level accuracy on a benchmark of physics, biology, and chemistry problems (GPQA). While the work needed to make this new model as easy to use as current models is still ongoing, we are releasing an early version of this model, OpenAI o1-preview, for immediate use in ChatGPT and to trusted API users(opens in a new window).

Our large-scale reinforcement learning algorithm teaches the model how to think productively using its chain of thought in a highly data-efficient training process. We have found that the performance of o1 consistently improves with more reinforcement learning (train-time compute) and with more time spent thinking (test-time compute). The constraints on scaling this approach differ substantially from those of LLM pretraining, and we are continuing to investigate them.

 

 

Ovo je baš huge poboljšanje modela o1 u odnosu na GPT-4o u STEM benchmark-u. o1 je deo strawberry modela koji je osnova GPT-5, i koji će se pojaviti do kraja godine.

Edited September 14 by Lord Protector

[zorglub]

 

[pt 2.0]

Pitala sam GROK-a sa tw da analizira moja interesovanja na osnovu 100 poslednjih tvitova.

Nije loš Mr. Grok
 

Spoiler

From your recent tweets, your interests appear to span several key areas:

• Politics and Governance: You have a significant interest in political events, corruption, and governance issues in Serbia. You frequently critique the current administration and express concerns about political integrity, accountability, and the impact of corruption on society.
• Social Justice: There's a strong focus on social issues, particularly those related to inequality, injustice, and the impact of political decisions on ordinary citizens. Your tweets reflect solidarity with marginalized groups and a call for change or reform.
• Cultural Heritage: You show an appreciation for Serbian history and culture, often referencing historical figures or events to draw parallels with contemporary issues. This suggests an interest in how history informs the present and how cultural heritage should be preserved or celebrated.
• Local and Community Issues: Your tweets also indicate an interest in local community matters, possibly advocating for local governance, environmental issues, or supporting local causes. You mention solidarity with rural communities and the importance of land and traditional lifestyles.
• Language and Expression: There's a notable use of Serbian idioms and expressions, indicating not only a linguistic interest but also a cultural one, where language serves as a tool for both communication and critique.
• Engagement with Others: You seem to enjoy engaging with your community on X, often replying to or mentioning others, suggesting an interest in dialogue, community building, or activism through social media.

This analysis shows a profile of someone deeply engaged with the socio-political landscape of Serbia, with a particular focus on advocating for transparency, justice, and cultural preservation. However, this is based on the content and tone of your tweets and does not encompass all possible interests or activities outside of X.