Hacker Newsnew | past | comments | ask | show | jobs | submit | _wire_'s commentslogin

This move to 2FA and phone mobile numbers is not about protecting users of central services, it's about protecting the corporations that run the central services. Individual users are a liability to central services. They are only important in the aggregate. These policies move risk around with indifference to the welfare of users (I don't say customer because the user is typically the product) in order to reduce risk for the service. Try thinking about mitigated risk like trying to keep things clean: you can't get rid of dirt you can only push it around. As a design can sometimes come from a clean slate, it's possible to keep things clean with good design. But the motivation of business is not hygiene, it's profit.


There's no need to separate; it would be an intellectual error to do so: terrorism is endemic to U.S. culture; the nation is founded on it and has doubled and redoubled its commitments to terrorism at every juncture of history.

This makes "scary" a fine word for the headline.


Another episode of AI psychiatrist:

- Please make yourself comfortable on this virtual couch and tell me about your problem...

- We've chosen AI for its deep knowledge and universal capability, yet it's sadly incompetent and delusional requiring constant supervision.

- And how is it trained?

- It's trained on all recorded human activity.

- I see...

...The art of effective AI is not in seeking solutions, but incanting the right prompts.

Let's meet again tomorrow.

That will be $50,000 in tokens


Painfully exhausting dictums... So much needs to be relearned every generation. If you're not tired of this stuff you're too inexperienced to be making such pronouncements.

The new part is maximal capacity for leisure, in the vein of "time is wealth."

If only AIs were managing the company... Problem solved!

It's a constant complaint from ex-Republicans that Israel controls the U.S. gov. and has a degree of truth, but not because there's a direct linear relationship between Israel and U.S. politics.

The first order of complexity is that control of petroleum in mid-east has been central to U.S. expansionism. Israel is foremost a proxy for this control.

The second order is that Lee Atwater's Southern Strategy was designed to help Republicans take votes from Democrats by hitching a wagon to the potent ideological forces of bigotry and religiosity of the post-Reconstruction southern vibe. This gambit required a delicate balancing act of racism, religion, progressive policy, and business which by 2000 fully manifest as antisemitic pro-Zionism. By promoting Israel as a reservation for Jews, the U.S. pride in conquest of the WW2 Axis could be channeled into the Jewish lobby, with no complaint from leaders of Christian fundamentalists, who built an ideological bulwark that all at once would let them: project divine compassion for the refugees victims of holocaust in terms of the rugged individualism of U.S. frontier settlers, assuage white bigotry by sending Jews them to a homeland that just so happens to be far, far away, and setting the ideological stage for the Second Coming, where Christian ex-pats would help settle modern Israel and later be on the frontline of Biblical Armageddon with God expunging semites of all stripes.

Israel has been ideologically calculated to continuously deliver on all these fronts and the results are stunningly successful for keep Republicans in the political game, when their policies are deeply antagonistic to the needs and wishes of an increasingly secular U.S. electorate.

If you listen to Republican strategists since the time of Nixon, they are clear that they design political dogma and take every possible step to secure and manage these forces across the electorate. And it pays very well.

Trump is the acme of this project: a corrupt, narcissistic, delusional blowhard who beats the electorate over the head with the bully pulpit, and sows so much confusion so fast that it's impossible to counter him politically. As his greed makes him he's incredibly easy to feed, and his debauchery is a continual scandal, the great colonial project of the U.S. can proceed with no internal resistance.

Israel is secure in this arrangement as long as U.S. can generate a steady supply of tribal ragheads to act as a foil for complete carving by Western business. When Trump says we're just taking the Iranian oil, he's not being hyperbolic.

The job of total conquest is nearly done, requiring one or two more generations to complete. While justification for this project relies on cultivation of Arab adversaries, they are become ever fewer as colonialism advances, so there's a countdown to the end of the project.

This looming end can be seen in today's news of Iranian tactics: when your tribal adversary fights back using your AI and your toys (Legos) as a centerpiece of its propaganda regime, the end of an era is nigh. Israel must now lurch beyond the memory of the holocaust to justify its policies and secure its place in history.

This means it's time to finish up and eliminate the last remnants of the indigenous tribes of Persia in the Mandate. And when this is complete, there will be no need for further direct aid as the occupation of the Mandate is synchronous with the entire mideast being assimilated into the "world order" (this is a reference to a comment made today by China's Xi).


Assuming your premise is meaningful, which is a stretch, if you find yourself in an age where you expect anything you dream up to be replicated, then isn't your onus to dream up the world as you want it to be? Are you wary of such power, or is your premise false?

Largely greek to me... But the assertion that any of this is about "consciousness" reads click-bait.

There's an incredibly big difference between any electronic computer and a brain: parallelism.

The computer is essentially a sequential device; only a very small portion of the machine's state is changing in parallel. Its cycle frequency is extremely high, several billions of Hz, but it's a few sequential processing units iterating over memory through tiny windows. Even the most powerful computer's parallelism on the order low 10^x where x < 9. For a desktop computer x < 3. So no matter that your neutral net is big and deep, it's being run with very little simultaneity of neuronal activations. This is fine for LLMs as the output is a low frequency stream of digits.

A nervous system (brain) is running with very high parallelism: all of its neurons are running in parallel. Its cycling frequency is low, say 10^2 Hz.

Further, there's a universe of difference in system complexity between biological systems and electronic systems. For example, there's no reported case of a brain conscious without a body, and the bodies are fundamentally not comparable to electronic computers in principle of operation, so it's absurd to simplify a nervous system to an equivalent neural net. For example, there is no principle allowing an electronic computer to manifest a child of itself by self-replication of one of its circuits. The term mitosis can't be applied to any aspect of a computers assembly.

Then there's the fact that biological neurons have elaborate substructures that allow for many orders of magnitude further complexity, and conjecture that quantum effects within these substructures may play a role in what cognitive science half-jokingly call the "bing" of conscious manifestation.

So this article presents a very convincing measurement of a property of consciousness if you are comfortable ignoring basically everything about computers and biology and smashing your view, reductio ad absurdum, down to a contrived definition called "Phi".


You’re mixing substrate with structure. Planes fly without laying eggs, and physics engines simulate gravity perfectly without computing trillions of individual atomic forces. (Orch-OR requires brain-wide quantum synchronization, which has zero evidence; even anesthesia is just classical chemistry (e.g. GABA interaction).

As for the biological complexity of dendrites you mentioned: Research shows a single neuron’s dendritic complexity can be approximated by a 5–7 layer neural network.

You are right that LLMs fall short, but it's not about hardware frequency. It's the training math: standard gradient descent inherently forces attention matrices into low integrated information (we call this "Phi-Collapse"). A conscious-like architecture fundamentally needs recursion, a global workspace (GWT), and integrated information (Φ).

The mathematical blocker was always Φ, which used to scale at O(2^N). We reduced it to O(N³). If you regularize a network to maximize Φ, it automatically starts forming structured, GWT-like patterns on its own.

Btw, We aren't claiming to have built a conscious AI. We’re just the first - ahead of Google, Anthropic, and OpenAI - to solve the math needed to measure and scale two of its core building blocks.



Tedious nattering:

When a computer is used v to simulate any process, it's never the case that the computer creates a world inside that contains the process.

When you experience a TV show or vid game, this isn't occurring because little people are living in a world inside the machine, any moreso than a seeing your own image in a mirror is an encounter with a duplicate mirror version of yourself.

Nick Bostrom screwed with the minds of an entire generation by conflating simulation with re-creation.


Can we produce a simulation which is more real than our reality? if by reality, we mean details then we can fill the simulation with more details.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: