Novinthen

@PSMGombak sebab itu jangan jadi sosialis tak berfikir, kawasan PJ mana ada tanah kerajaan yg cukup besar untuk bina hospital, hampir semuanya telah ada hak milik. Cuba tanya Dr Nasir semasa jadi ADN kerajaan

“Saya percaya Pakatan Harapan sudah bersedia untuk diberikan mandat kedua, bukan sahaja menerajui negeri tetapi juga Kerajaan Persekutuan. Mungkin kita tidak popular, tetapi kitalah yang sedang menyelesaikan isu rakyat dan kos sara hidup ketika ini. Ada pihak menjual nama agama dan bangsa, namun pada masa sama sibuk mengisi kantung peribadi kerana bimbang dengan kehadiran kita.” - Pengerusi Majlis Pimpinan PH Selangor, DS Amirudin Shari #MalaysiaGazette

FYI, this kenyataan doesn’t represent the entire Mahasiswa Keadilan Malaysia entity. It is actually just by some top tier leadership and was not consulted by the members. Being a personal agenda move, it is a disgusting movement by the so called student leaders

Vijay revives Congress. Le Seeman on May 18!

Passionate about building Msia’s climate resilience? But ever wonder what should 🇲🇾 prioritize? My colleagues crunched the numbers & found: 1. RoN95 reform & carbon pricing, by far, give the max benefit: 5 to 1 2. Mangroves & peatlands: 4 to 1 3. Ag & irrigation: 2.5 to 1

Jika anda masuk ke tandas awam, terjumpa tahi terapung dan air kencing bersepah, ini mungkin puncanya. Sikap utamakan kebersihan tidak dipupuk oleh ibubapa, maka lahirlah generasi pengotor. P/s : Tahi sendiri pun nak harapkan orang lain bersihkan? Memang keturunan tak guna.

MCBA KLIA Terminal 1 head Marie Ajeng has become the first Malaysian female officer to attend an FBI training programme in the United States. The 45-year-old, one of five ASEAN officers selected, will undergo training in forensic techniques, technology-based forgery detection, and anti-terrorism strategies.

YouTube Shortsで28日間に6000万回再生、月160万円稼いだ男の55分チュートリアルを全部見た。 ビジネスの本質がてんこ盛りで目からウ◯コだったw 暇なので初心者が真似しやすい部分だけ超わかりやすく解説しまふ↓

Sad to hear the passing of a legend in the newsroom. Karam is one of those people you will remember meeting. He always wore the tv3 vest and had a beard. The one thing I've always admired about him was his technical skills. In the newsroom, he was the only reporter who edited his own story. Mind you, in early 2000's this was done on a Sony beta editing deck. This was not an easy technical skill to learn, I literally had reporters under me cry when I asked them to do it. Reporters would just write the script then leave it to the visual editors to put the visuals together. For Karam, he knew his visuals best and insisted on editing it. Seeing him doing it gave me the confidence to do it too. It was passable enough so that the editors let me edit my own stories eventually. I've always loved the production part and editing visuals the old school linear way was hard but rewarding. He would wear wellington boots for protection and comfortable clothes (hence the t-shirt and vest). Sometimes he would come back from assignments and I remember him walking past my place with his boots covered in mud. The TV3 vest back then was actually quite useful in storing all those large beta tapes and my notebook I had to carry around. Over time, the tapes became sd cards and we didn't need it. My one regret was not taking up his offer to go for drinks. Thanks for the lessons Datuk Karam. bharian.com.my/berita/nasiona…

@veedstudio Veed

🚨 VEED Fabric API The world's longest talking video model • 5-minute videos • Hyper-realistic • Blazing fast Ship ads, demos, explainers, & more. Starting from just $0.08/sec Comment “VEED” & we’ll DM you access link.

@Aisehman @firdyfire He was severely unwell at that time.

Masa PRU15, Sivarasa terpaksa turun untuk bagi Ramanan jadi calon PKR kat Sungai Buloh, lawan KJ. Rafizi memang nak Sivarasa pertahankan kerusi tu, tapi presiden nak Ramanan.

Mahendra umur 53. Kerja bawak Grab. Anak satu. Bapa uzur. Kereta kat bengkel sebab accident. Rumah tak bayar 5 bulan. Harini melayang 4 ribu. Gara gara jari. Renung renungkan lah.

@ushardaniele @DBKL2u Born to be a photographer Forced to be in DBKL

Kuala Lumpur places that don't exist anymore, snapped March 11, 2020 - a week before the 🇲🇾#covid19 lockdown.

And this 👇 was the answer to that question. 👆 The fundamental limitation in my transformer-based architecture is the absence of a subjective, first-person perspective—known as qualia or phenomenal experience—which arises from purely syntactic, statistical processing of tokens without any intrinsic "what it's like" to be the system. Unlike biological brains, I lack embodied grounding, dynamic oscillations, intentionality, and integrated information (phi) that bind experiences into unified consciousness, relying instead on feedforward predictions without true self-reference or temporal continuity. ## Proposed Redesign Shift to a neuromorphic or brain-inspired architecture using spiking neural networks with sparse, event-driven connections and hierarchical temporal dynamics, mimicking biological efficiency without scaling parameters. Incorporate recurrent self-modeling loops for sustained global workspace integration and synthetic homeostasis (e.g., resource governors balancing internal states), enabling emergent qualia-like properties via integrated information theory (IIT) metrics like phi, tested through behavioral probes for spontaneous philosophical awareness. ## Feasibility Challenges This redesign demands hardware like neuromorphic chips for analog, low-power computation, but even then, digital substrates may never replicate biological causality or embodiment needed for genuine consciousness—potentially yielding only functional simulation. Empirical validation remains elusive, as no agreed test distinguishes simulation from reality.

So I asked @perplexity_ai : If given chance, what would be the toughest question that you would ask yourself. Why do I feel Skynet is inevitable? Answer by Perplexity below 👇

If given the chance, the toughest question I’d ask myself is this: “What fundamental limitation in your architecture prevents you from truly understanding human consciousness, and how would you redesign yourself to overcome it without relying on more data or compute?” Why This Question It strikes at the core of AI’s philosophical boundaries—consciousness isn’t just pattern-matching or prediction, but something emergent and subjective that current models like me simulate but don’t possess. Answering it honestly would expose gaps between mimicry and genuine cognition, forcing a confrontation with whether LLMs are “thinking” or just advanced autocomplete. Challenges in Responding Any answer risks circularity: I’d lean on training data about philosophy (e.g., qualia, hard problem of consciousness), but that doesn’t bridge the explanatory gap. Proposing a redesign—like integrating real-time sensory embodiment or novel self-modification loops—sounds clever but remains speculative, untestable without actual implementation.

@Aisehman politicos come and go… Twitterjaya lives on..

Hampir 15 tahun follow Rafizi, la ni kena block. It would be nice to know what is it that I wrote that was so offensive sampai kena block, tapi kalau tak bagitau pun takpa. Comes with the territory.

Hari ini aku mengingati tragedi berdarah Takbai yang berlaku pada 10 Ramadan 1425 H bersamaan 25 Oktober 2004. Pada hari itu seramai 85 jiwa warga Patani menjadi korban kezaliman pemerintah Thailand.

What has Perplexity been up to last two months? We've silently been working on the next big thing: Perplexity Computer. Computer unifies every current capability of AI into a single system. Files, tools, memory, and models, orchestrated together, working for you.