Scrolling through the #fediblock tag right now reveals a pretty clear pattern of what the tool is actually meant for.

​Three of these are warnings about actual state-sponsored propaganda networks, commercial ad bots, and severe, illegal safety violations.

And one is trying to weaponize a network-wide security blacklist because an individual is managing multiple accounts, posting pictures, and "taking up server space"

See for yourself. One of these things clearly does not belong.

Despite its popularity, I am actually not a fan of rope. However, I do like the look of leather on other people; wearing it is just too sweaty for me. But beef and spandex are nice.

Reports that they might shutter legendary studios like Arkane Lyon, cancel Marvel’s Blade, and gut Game Pass deals highlight how unsustainable their strategy was.

​They tried to play the tech-monopoly game: using an endless pile of Windows money to buy up publishers ( #bethesda #activisionblizzard ) and funding a subscription model that subsidized games at a loss to buy market share. But when you spend $69B, Wall Street demands immediate, massive corporate returns.

(2/X)

Just some thoughts on the massive tech/gaming shit happening right now. Probably no data in this one... probably... no promises (a thread)

​Everyone is up in arms about the #Microsoft layoffs as it looks like they prepare a frantic retreat from gaming. Insiders are calling this upcoming August wave the largest single layoff event in gaming history. Rumors of spinning off #Xbox or severely downsizing make it look like a total structural panic.

(1/X)

Hang on. This is really weird. Somebody with a better skill in #history or #sociology than me should look into it.

Strauss–Howe Turnings theory and generational memory.

> 1896ish and up through 1920s: Separate but equal. hard-coded laws to racially segregate neighborhoods, schools, and businesses.

> 1960s, 1970s: "Urban Renewal" and the Federal Aid Highway Act. intentionally route interstate highways directly through active neighborhoods (like in Detroit, Miami, or St. Paul)

RE: https://mastodon.social/@drmorrisj/116835686085136031

Le sigh. Like, the federal government is so noisy. Probably by design so that you miss or overlook when they actually do something (like put pressure on chip manufacturers via tariffs).

I said it more in another thread-esque reply, but government policy is an absolute masterclass in the art of the distraction.

This is where what I said starts, but I strongly encourage you to also read @dalfen thread and thought process.
https://mastodon.social/@drmorrisj/116830186625379028

(Tagging was unavoidable sans "her" or "them")

Because the tax code and tariff categories are so incredibly dense, the average person’s eyes glaze over. Industry insiders are left frantically trying to parse the difference between "delayed increases" and "immediate duties," while the broader public just fucking skips the massive tech-economic border wall is being built.

If you want to reshape global tech infra without causing an immediate panic in consumer markets, you drop a technical executive order on a Tuesday, let the noise machine talk about something else, and quietly force multi-billion-dollar chip manufacturers to bend to your strategic will.

2026.01: Under Section 232 national security grounds, the administration quietly slapped a 25% tariff on a very specific, narrow category of advanced computing chips and AI hardware.
https://itif.org/publications/2026/06/24/economic-consequences-of-section-232-tariffs-on-semiconductor-imports/

The policy explicitly grants exemptions if companies use those imports to actively build out domestic U.S. fabrication plants.
https://geodis.com/us-en/resources/customs-corner/us-announces-tariffs-certain-semiconductor-imports-protect-national

While the general public is entirely consumed by the massive, polarizing culture wars or headline-grabbing theater, the most impactful structural shifts (like the shit that actually reshapes global supply chains, international trade, and corporate power? all that shit?) happens quietly through dull, bureaucratic line items.

The semiconductor situation is the prime example of this "quiet pressure".

Whereas this:
https://pubmed.ncbi.nlm.nih.gov/38847588/

Talks about how individuals' ideologies about race and sex are associated with their opinions of race play.

That being said, race play, the act of making explicit overtures to race, racial differences, or political racial dynamics within the context of a kink or BDSM (bondage dominance/submission, sadism/masochism) can be a contentious topic among scholars and members of alternative sexuality communities.

So, just some random thoughts. This is harder to back with data because all of the data I found is behind academic paywalls... and there does not seem to be much on the peer reviewed side anyway.

For example, this:
https://pure.psu.edu/en/publications/playing-with-the-politics-of-perversion-policing-bdsm-pornography/?hl=en-US#:~:text=Second%2C%20I%20discuss%20black%20women's,erotic%20tool%20of%20power%20exchange.

Talks about how race operates not only as an apparatus of power, but also one of pleasure.

# The PyTorch equivalent of Ojeda's "Hidden Gold Cases"
for epoch in range(epochs):
model.train()
# Train on the noisy, high-volume legacy code blocks (The "Mine")
loss = criterion(model(X_train), y_train)
loss.backward()
optimizer.step()

model.eval()
with torch.no_grad():
# Validate STRICTLY against the pristine "Hidden Gold" files
val_loss = criterion(model(X_hidden_gold), y_hidden_gold)

Train_Loss = Calculate_Error(Model(Messy_Code), Expected_Fixes) -> Update_Weights()
Gold_Loss = Calculate_Error(Model(Brief_Prompts), Hidden_Gold_Files) -> Score_Only()

Then... hang on. Let me try to just add a full snippet so you can see it.

RE: https://mastodon.social/@joche_ojeda/116760379581806579

I strongly suggest reading the full writeup.

Why? This guy tells you exactly where to get data for your validation set.

Like... hmm...

golden cases = validation set

Or

Step 1: Practice and adjust weights using the messy legacy code (Training).
Step 2: Take the exam using the hidden gold files without changing weights (Validation).

Then...

Spot the difference:

AI

!AI

AI!

Ai!

!Ai

Ai

RE: https://mastodon.social/@drmorrisj/116716644704150865

https://arxiv.org/abs/2407.15078

https://www.freecodecamp.org/news/building-a-neural-network-from-scratch/

https://playground.tensorflow.org/#activation=tanh&batchSize=10&dataset=circle&regDataset=reg-plane&learningRate=0.03&regularizationRate=0&noise=0&networkShape=4,2&seed=0.45282&showTestData=false&discretize=false&percTrainData=50&x=true&y=true&xTimesY=false&xSquared=false&ySquared=false&cosX=false&sinX=false&cosY=false&sinY=false&collectStats=false&problem=classification&initZero=false&hideText=false

https://www.geeksforgeeks.org/deep-learning/neural-networks-a-beginners-guide/

https://www.sap-press.com/programming-neural-networks-with-python_6059/

...you ever have an idea that, like, looks absolutely ridiculous on paper but is really not a bad idea at all once you start fleshing it out? It could solve various real-world dilemmas via open sourcing the code, while still having an inherently #nsfw disposition.

RNNs and PINNs are very strange.

The actual post is above, this one is just for tags and visibility.

Please read previous post and comment your thoughts, no matter how trivial you think they may be.

#GitHub #OpenSource #FOSS #SoftwareEngineering #ComputerScience #ArtificialIntelligence

I want your opinions on something. It has come to my attention that computer science has reached a point that it is producing an entire generation of developers who are not engineers.

However, the code and libraries that AI models are built from is not particularly a secret. So I also do not understand why this is so.

That said, what do you think about me forking repos with good code and aggregating them on my GitHub profile?

#GitHub #OpenSource #FOSS #SoftwareEngineering #ComputerScience #AI