Prgnr

A PhD scholar studying NLP and automation of text processing. I used to do art & animation before generative AI was big.

[Prgnr @MovingFramesP]

@ThrillaRilla369 Gorilla

[Prgnr @MovingFramesP]

In abundance, some thrive materialistically while some experience abyss and beyond.

[Prgnr @MovingFramesP]

Most often we have only a single perspective on a case that we hold as true.

Instead, it is far better for each of us to accept that our truth is only a single possibility out of many.

Objectivity of truth only exists in science. Otherwise, it's only perspectives and interpretations.

It is not offensive and it is very much okay to have our personal truths. However, this small shift in our mindset opens up a multitude of cognitive layers.

[Simon Maechling @simonmaechling]

The problem isn’t that people are stupid.

The problem is that many people don’t realize they are being psychologically manipulated by systems designed to maximize engagement.

[Prgnr @MovingFramesP]

@Strife212 Interesting. If someday it turns out that what we think and say is computable manually, will it qualify us as conscious?

[𐌁𐌉Ᏽ 𐌕𐌉𐌌𐌉 @OrevaZSN]

Libraries: free.
AI: subscription.

Libraries: written by humans with expertise.
AI: trained on whatever was on the internet.

Libraries: staffed by professionals.
AI: confidently wrong.

Go to the library.

[THE SKIN DOCTOR @theskindoctor13]

The govt should consider creating a national pool of young ethical hackers, cybersecurity researchers, and talented students to regularly test the security of its websites and digital infrastructure.

Every year, crores are spent on private contractors to build and maintain govt portals. Yet many remain slow, poorly designed, vulnerable to security flaws, plagued by glitches, and raise concerns about data privacy.

As the recent CBSE controversy showed, even a small technical vulnerability can trigger confusion, mistrust, and nationwide outrage. So why not tap into the talent that already exists in the country?

A structured bug bounty and security-audit program, with attractive rewards, would encourage some of India's brightest minds to identify vulnerabilities before malicious actors do. The cost would be a fraction of current spending and should be made part of contractors' obligations.

While some govt platforms already have bug bounty programs, the current approach is either not comprehensive enough or not effective enough. A stronger, centralized, and better-funded system may be needed. No system is perfectly secure, but having thousands of skilled people actively looking for weaknesses is more effective than relying solely on a handful of vendors.

[Mo @atmoio]

I found the actual reason AGI isn't real

[𐌁𐌉Ᏽ 𐌕𐌉𐌌𐌉 @OrevaZSN]

App idea: Shazam but for the sound my car is making.

[Emma @EmmaDarles]

En tant qu’enseignante-chercheuse en informatique, je suis toujours surprise par la pauvreté du débat sur l’IA.

D’un côté, on nous explique que c’est une force obscure que personne ne maîtrise et qu’il faudrait arrêter immédiatement. De l’autre, qu’elle va résoudre tous nos problèmes.

Les deux ont tort.

Non, l’IA n’est pas une magie incontrôlable.
Nous connaissons les architectures que nous développons.
Nous savons comment elles sont entraînées.
Nous mesurons leurs performances.
Nous savons les améliorer.

Le vrai sujet n’est pas la maîtrise technique. Le vrai sujet est l’explicabilité.
Oui, il existe aujourd’hui de réels progrès en intelligence artificielle explicable (XAI).
Mais même avec ces avancées, un modèle peut être extrêmement performant tout en restant incapable d’expliquer de façon claire, fiable et complète chacune de ses décisions, en particulier lorsqu'il s'agit des modèles les plus complexes.

L’IA traduit plus vite. Analyse plus vite. Classe plus vite.
Très bien.
Mais parce qu’elle va plus vite, devons-nous lui laisser décider à notre place ?

Déléguer une traduction n’est pas déléguer un diagnostic médical.
Déléguer une recherche documentaire n’est pas déléguer une décision de justice.
Déléguer une assistance au pilotage n’est pas déléguer l’usage de la force.

La question du XXIe siècle n’est pas : « L’IA est-elle dangereuse ? »
La question est : Jusqu’où sommes-nous prêts à déléguer notre jugement à des systèmes qui apprennent du passé, reproduisent parfois ses biais et ne portent jamais la responsabilité de leurs décisions ?

[KeriA @KeriA1776again]

Paywall be damned. Work around to academic studies behind a paywall. Clever gal.

Source: https://t.co/X8lh4cXbv8

[Prof. Carl Sagan @ProfCarlSagan]

If we don’t explain science to the public, others will fill the gap with nonsense.

[𐌁𐌉Ᏽ 𐌕𐌉𐌌𐌉 @OrevaZSN]

In case you forgot, we live on a planet where lightning turns sand into glass, dolphins give each other names, the ocean glows in the dark, and rain has a scent before it falls. You stand here with a beating heart, a steady moon above, and a world quietly, absurdly beautiful.

[Prgnr @MovingFramesP]

Bard flies away after scratching and chirp chirp

[Abhishek Singh @abhiart]

It’s extremely concerning what we in India are doing to our rich ecology, this mindless, unregulated development for money, has become dangerous to living conditions. We must integrate climate design into our plans to say the least or else, this indifference will be our undoing.

[Pope Leo XIV @Pontifex]

Deep inner suffering inevitably arises when the human person is reduced to performance, consumption, or a statistical datum. Many young people today live under the yoke of expectations to perform, immersed in an exasperated competitiveness that generates anxiety, fear of not measuring up, and disorientation.

[Prgnr @MovingFramesP]

I have become aware of at what point the attention span of the person i am speaking to is going to expire.

[Prgnr @MovingFramesP]

If we consider the possibility that everything that we experience can teach us something, our brain adapts to this habit. World becomes more vibrant and you are never bored.

[Prgnr @MovingFramesP]

@SadhguruJV It should become a rule to plant >1 tree (and take care of it) per tree destroyed by a human (org/etc.)

[Prgnr @MovingFramesP]

We all should know that we know nothing. So what is the next step after we realise that we know nothing? We put efforts to know atleast something whenever it is possible. Because the knowledge of few something stack monotonically positively. Unlike in AI that stack tokens.