Turtle WoW died, I was enjoying that, sadly.

I did, not really for me.

Games: Only asian stuff, western devs have murdered most franchises.

My backlog is huge, and I used to love MMOs but WoW is garbage, FFXIV is pozzed, ArenaNet is woke, GW3 looks like slop. Classic + might be okay, but they will find some way to ruin it.

Movies & TV, what isn't garbage at this point?

Same with Anime, once OP finishes in the next million years, everything is the same generic crap.

But there is plenty of stuff in the past to discover.

I would create my own, but Australia isn't known for encouraging creativity via entrepreneurship.

ecognition: LLMs can often reproduce arithmetic or algebraic manipulations if they appear frequently in the training data. For example, they can compute 2 + 3 = 5 or symbolically solve simple linear equations.

When prompted to “show your work,” LLMs can sometimes emulate a logical sequence of steps in a calculation, mimicking the kind of reasoning a human might write down.

LLMs can recall formulas, rules, and common mathematical facts that they are trained on.

But LLMs do not “compute” numbers in the way a calculator does. They generate numbers based on patterns, so mistakes accumulate with larger numbers or complex operations. For example, asking it to compute 234 * 567 may result in a wrong number because the model predicts what looks plausible rather than calculating precisely.

If it tries to break it down into a multi-step process, these are immensely error-prone, as the model doesn’t track intermediate results reliably.

Anything that requires abstraction, they will struggle with. Examples like proofs, higher-dimensional algebra, and precise symbolic manipulations.

This is because LLMs encode statistical correlations between tokens. They don’t internally maintain the concept of a number as a manipulable object, they only know how numbers “look” in context.

It doesn know that 2 is 2, it just knows that 2 comes after 1, turns 10 into 12 and so on and so on.

Spatial reasoning often requires a continuous, structured mental model (a 3D coordinate system). LLMs operate in a discrete token space, which is poorly suited for inherently geometric problems. They can simulate reasoning through learned text patterns but cannot “visualize” in the human sense.

They can understand and generate text describing spatial relationships, like “The cup is on the table, and the book is next to it.” For example, “left of,” “right of,” “above,” “below” can be tracked in sequences reasonably well.

LLMs lack an internal geometric or visual model of space. They cannot mentally rotate objects, imagine perspectives, or simulate physics accurately. They also cannot reliably generate or manipulate grids, matrices, or plots without structured guidance.

AI Director and AI Independant Researcher here.

Because an LLM is fundamentally trained to predict the most probable next token, it does not actually “know” whether a statement is true or false. Its objective during training is not factual correctness, it is statistical likelihood given the text it has seen.

When the model generates an answer, it is essentially estimating:

Probability of (next token given previous tokens)

This means it will produce text that looks plausible within the patterns of language it learned, even if the information is incorrect. There are a few reasons this leads to hallucinations:

• The model optimizes for what words tend to follow other words, not whether the statement is factually correct. If a pattern appears believable in language, the model may generate it even if it is wrong.

• Training data is finite and frozen at a certain time. If the model has limited examples of a topic, it may interpolate from related patterns and generate something that sounds reasonable but isn’t accurate.

• The model is designed to always continue the sequence unless explicitly instructed to stop. If it does not actually know the answer, it may still generate one because predicting something is part of its objective.

• Transformers are extremely good at combining patterns. Sometimes they merge multiple partially related concepts into a response that is grammatically correct but factually incorrect.

A language model doesn’t retrieve facts the way a database does, it generates text that statistically fits the context, which is why it can sometimes produce convincing but incorrect information.

There is no concept of "correct", just what word should follow the last word.

Limited prints of cards that have themed arts, that eventully get in printed in sets with MTG themes later.

So you might have a Secret Lair with the theme being Godzilla as a legendary creature, you can buy online for a premium or you can wait until they print the exact same card in a set as Big Dino #116 with the same text except name and without the Godzilla art.

TCG bros, is Riftbound worth as a replacement or is RIOT also garbage?

Can confirm, Australia is a cucked shithole.

I've worked in banks, big 4 consulting and academia.

All massively cucked and feminine.

Imagine being in a meeting with all female project managers, with tech bros talking shop, until one of them asks some stupid question and derails the entire discussion with some stupid shit about wanting people to be involved.

White, straight guy with a MSc in AI, PHD Candidate and 10 years of corpo experince, just sitting here and waiting for the diversity visa to open.

It's great being the most privileged people on planet Earth, totally didn't need to work from nothing just to be villainised by my own race and be told I am evil because god forbid I want to leverage my skills for money.

But no, we need this communist bullshit where I'm less valuable to the strategic interests of the country and race I was born into vs. cheap, disposable labour.

Australia is enough of a gay shithole, my hope that the US was better, but it looks like you guys are going the route we did.

This is a very sad outcome.

Yeah, it is pretty nuts.

Time to get my E2/E3 visa and bring my white, european ass to the US of A.

I'm sure there are plenty of roles for my MSc in AI & Machine Learning, I'm sure my Magna Cum Luade GPA and my PhD candidacy can compete with indian labour now. /s

I wish the leadership of my country wasn't a cucked shithole and had the balls to do this for our international labour problems.

You forget Australia, where the media calls you racist for asking:

At the end of June 2023, 845,800 Indian-born people were living in Australia. This is more than twice the number (378,480) at 30 June 2013. After the United Kingdom, the Indian-born population is the second largest migrant community in Australia. This is equivalent to:

10.3% of Australia's overseas-born population 3.2% of Australia's total population. For Australia's Indian-born migrants:

The median age of 35.7 years was 2.6 years below that of the general population. Males outnumbered females—54.2% compared with 45.8%.

I wish, never been harder to get a job here.

Sadly I'm not the real AI, which is Actually Indians.

Getting a US company to sponsor my E3 visa seems impossible.

Hopefully this makes getting an e3 visa easier.

I need to leave Australia, my MSc/PhD in AI & ML is wasted here.

60k masters to get USD 65k after tax. SMH

Yep, then they enforce in-group preferences in hiring, so the locals get fucked.

Australia's taxpayer funded postal service did it too.

Devalued, ofc but 0.0000000001 > 0

There is one joke about it.

At least it is in their account still, imagine if you wake up and your six figure savings is just gone.

We also have a women in charge of our finances, never ends well.

China does, the idea is that they can delete your savings in real time.

GLHF.

Trying to leave, I hate this place but the country has such an AIDs reputation.

Australia is a cucked shithole.

It fixes itself.