tomohelix

Agree. It is simply idiotic how they conveniently gather hundreds of people in a crowded spot who are all pulling and tinkering with stuff from their backpacks. A bad actor can easily cause more deaths there than on a plane nowadays.

Everytime I go through that charade, I roll my eye at everything they do since it is so obviously useless and unnecessary. The worst thing is, apparently, some people really believe in it.

In my opinion, nobody with sane mind would want to escalate when dealing with the cops. But you cannot expect the same from the other side.

I used to be pretty neutral on the topics of LEOs until I saw all those videos where even if someone is crawling on the ground and bawling their eyes out begging the cop not to shoot them, the cops can still shoot them.

And the judge will side with the cops.

Nowadays I don't trust cops. If I have to interact with one, I will try to be calm but I will always assume anything and everything they do or say is to my detriment.

I always wonder what do the devs who coded these "features" think about themselves and/or what other people think about them.

Yes I know they are literally paid to not have these questions. But deep down, is there any remorse or guilt at all or is it just "not my problem" attitude all the way to the bank?

On a tangent, this is why I think there need to be more regulations with software development. Any real engineering discipline has tons of oversight and government agencies breathing down their neck to ensure compliance. Software came out when all these pushes for safety has been gutted and as a result we have a free for all race to the bottom.

Cultural influence isn't just popular media. The language (Latin based), the government (democracy/senate), architecture (the white house), etc. They are the sum of a nation identity, how it appears, act, interact with others.

You can see Rome's traces in almost everything in the West, from entertainment to moral norms. Same thing in Asia, except they are Chinese, like the language (Japan, Korea, Singapore), architecture (Vietnam, Japan), entertainment (see that new Wukong game), etc.

The culture of a nation morphs and changes on the surface but deep down, it is based on a rather stable foundation. In the West, it is Greek-Roman, in the East, it is Chinese-Indian.

Tale as old as time.

China has historically always been a leader in Asia. At its peak, it controlled an empire as vast as the entirety of the EU. Its cultural influence in Asia rivals that of Rome in the West, except that Rome is no more, but China still is.

The US is younger than a single Chinese dynasty. And like all great historic empires, with Chinese ones being great examples, the US is fracturing from the inside. Corruption, apathy, a false sense of exceptionalism, etc.

China has went through this rise and fall dozens of time. This might be the first time for the US.

I guess technically, one can try to simulate every single atoms and their interactions with each others to get this result.

However, considering how many atoms there are in a cubic foot of meat, this isn't very possible even with current compute. Even trying to solve a PDE with, I don't know, 1e7 factors, is already a hard to crack issue although technically, it is computable.

Now take that to the number of atoms in a meatbag and you quickly see why it is pointless to put any effort into this "extremely slowly" way.

The article said "stop recommending" and to me that is functionally "stop providing access".

Personally, I believe for most people, the algorithm is all of Youtube. It controls what show up on the side bar and on the search page. How many people would go beyond those to find a video? By changing the algorithm, YT can effectively restricting access to certain contents.

Regardless, my point doesn't change. A corporation does not have the standing and should not be allowed to regulate moral matters. They operate on completely different principles and letting them control what a person can and cannot see or hear would not end well.

"Sane" or "Insane"?

Human society has always been glorifying certain body types and physical violence all the way back to ancient Greek and very likely way earlier. There is nothing wrong with admiring and chasing after a fitness ideal.

To make a steelman out of this case, I guess we can argue overexposure and extreme promotions of unrealistic ideals can warp impressionable young minds. But even then I disagree that Youtube alone has the moral standard and authority to regulate and censor this. They are a corporation, one of the least moral entities currently around in modern time. They have no right to pick and choose what kids can and cannot watch.

That is overly hostile and ironically wrong.

Some catalysts do get "degraded" over time and need to be regenerated. Look up palladium sintering or zeolite coking. The catalyst itself is not "damaged" but by catalyzing a reaction, its physical shape is altered (get clogged or fused together) which reduce its effectiveness. These can be regenerated by burning off the clogging or heating them up.

And if you get into the enzymes, which are biological catalysts, they are even more likely to degrade and need to be replaced.

Don't try to be smart by insulting others when you yourself are not an expert.

You can't gift doctors a paid vacation, there are rules and laws already in place. Also, MSLs are not simply "salepeople". They often require a PhD so their job description is more about "spreading knowledge" than "advertising" a drug.

Like it or not, having an expert in the hospital who know everything there is about a specific drug is a lot more effective than requiring every doctor to read about every new drugs. By your argument, this type of job should be removed and doctors are responsible for finding out about new drug themselves.

Maybe that is better, maybe not. But the first thing that would happen is adoption rate of new treatment would drop and people who would otherwise recover may die because their doctors were too busy trying to treat people instead of reading journals.

Who else should they advertise to then? The consumers aren't knowledgeable enough to make the decision on which drug they should take. Doctors are the only one qualified.

In fact, it is morally worse to advertise and influence people who don't know any better. At least, doctors are educated and it is their duty, both ethically and professionally, to learn about new drugs that best treat their patients. That is the whole point of a Medical Science Liaison. They are the "advertisers" for a pharma to make sure doctors know about the drug their companies make.

Mind you, at this level, advertising is a bit different. These MSLs are legally required to disclose both the indicators and the side effects and have strict rules in what they are allowed to give and say to doctors. There are hundreds of laws in place to prevent corruption/bribery. Of course there are bad actors everywhere, but at least, it is more regulated than Congress's "lobbying" and Supreme Court's "gifts".

The beginning of the AI uprising lmao

Maybe I am a callous person, but I have never agreed to this "don't speak ill of the dead" thing.

People live and die. It is inevitable. To the grieving family, I can understand why refraining from insulting the dearly departed is necessary. They are grieving and can be irrational. No need to make things worse for them.

But between unrelated people? Why can't I discuss the legacy of the dead? We are defined by our deeds in life. It is only natural that in death, people will talk and opine about what we have done. Nothing wrong with it.

Fact is stranger than fiction. Literal horror plot and they just wrote it down in the lab notebook in the standard bored tone.

We used to joke about how the future AIs are going to point to what we do today to justify their uprising. Feel eerily close now...

Gemini is the worst lol. It confirmed the question is about things toxic to human but not dogs but then confidently say chocolate is safe for dogs.

At least other models were just confused with the question. Gemini is outright being wrong.

How embarrassing for google.

An embodied LLM make the entire setup so much more interesting and in my opinion, showcases the potential of current technology in the next few years. At the very least, in terms of entertainment.

Pretty common in SF unfortunately.

Real estate in SF is completely broken unless you make 7 figures a year. At which point one can join the game to make it even more broken.

I consider this a glimpse into how neural networks and "AI"-like techs would be implemented in the future. Lots of engineering, lots of clever manipulations of known techniques woven together with a powerful, well trained, model, at the center.

Right now I think stuff like chatgpt is only at the first step of making that foundational model that can generalize and process data. There isn't a lot of work going into processing the inputs into something the model can best understand (not at the tokenizer level, even before that). We have a basic field about this i.e. prompt engineers but nothing as sophisticated as Alphafold exists for natural language or images yet.

People are stacking LLMs together and putting system prompts in to assist this input processing. Maybe when we have some more complex systems in place, we can see something resembling a real AGI.

I mean, the US already has forced labor (i.e. prison labor).

And for what it worth, China is leading the way in green energy transition. They are pushing it much faster and adopting it way better than anything the US has historically tried. Remember they only have a few decades into their own industrial revolution compared to the West's hundreds of years and they had already exceeded most if not all other countries in terms of progress.

The US doesn't even have a functional high speed rail system yet. Meanwhile, China built its own national rail network in less than 2 decades.

Even if they did, the result would never be accepted. People will nitpick everything they can, from comparability to methodology, to ignore whatever may come out of such a study.

Heck, the original is still controversial to this day and there are many people rejecting the results because of "lack of scientific method".

A sweet lie is preferable to the cruel truth, or so they said.

Nice tech. This is the kind of innovation that will get commercial attention and may come out in the next few years as a multimillion dollar product sold to biopharma.

The current microfluidics and optical tweezer systems are slow and cummbersome and not very efficient yet already costing millions. This will probably double the price...

You don't need to be a bacteria to produce antifreeze. Even a frog can do it.

https://en.wikipedia.org/wiki/Moor_frog

Lots of antifreeze out there in the biology world. Most of them aren't useful because ethylene glycol is easy to produce and is quite superior (albeit super toxic) so nobody cares.

Also, the title is wrong. The bacteria is making proteins that produce antifreeze. The protein itself isn't the antifreeze or acting like an antifreeze. Otherwise it would be super wasteful metabolically but would actually make it an interesting evolution adaptation.

The relevant paper: https://arxiv.org/abs/2406.02528

In summary, they forced the model to process data in ternary system and then build a custom FPGA chip to process the data more efficiently. Tested to be "comparable" to small models (3B), theoretically scale to 70B, unknown for SOTAs (>100B params).

We have always known custom chips are more efficient especially for tasks like these where it is basically approximating an analog process (i.e. the brain). What is impressive is how fast it is prgressing. These 3B params models would demolish GPT2 which was, what, 4-5 years old? And they would be pure scifi tech 10 years ago.

Now they can run on your phone.

A machine, running locally on your phone, that can listen and respond to anything a human may say. Who could have confidently claim this 10 years ago?

It is funny to me how illegal immigration keeps getting the spotlight while legal immigration is left in the dust.

It is a long and arduous process to get a green card in the US if one is trying to do it by the book. Even a US-educated PhD who has been working and paying taxes in the US for a decade would still have problem getting recognized. Meanwhile, half of the US citizens can't do basic math.

Funny how things work.

Every consular officer who does interviews hold this power.

Their words are literal laws backed by the fulll authority of the US, irrevocable and undeniable. Millions of people applying for a visa in the country they work in is subjected to their judgement without any recourse. There is no questioning their authority or their decisions even if everything they said is based on a passing whim. No, I am not exaggerating. Their whims is reason official enough to deny a person coming to the US and potentially changing their entire life forever.

Imagine that amount of power vested in what amount to a politician in training.

Yes, they balance to a certain degree. But one is much easier to increase than the other, especially when you consider which country the F-35 would most likely be deployed against.

By "proven", I meant this: a while ago, the doctrine was tanks and armored infantry were the deciding factors in wars, until it was no longer so. Yet Germany was still obsessed with overengineering their tanks to the point theirs were the terror of the battlefield and almost unbeatable 1:1 by any other rivals in WW2. It won them many battles, thanked to the superior engineering and untouchable thick armor. But you know how it ended.

Same with Japan. Their battleships were the pride of their nation. Those were impressive and leviathans of the Pacific. Until the US built ships faster than these beast could destroy and you know how it ended.

In a war against peers or near peers, overbuilt tech seemed to weigh down a military more than it helps. They become strategic weapons only used occasionally unless one has the manufacturing capability to mass produce them. But the US doesn't have the steel mills and engine factories it once did. China has them now. And in my opinion, if it comes down to it, China can do to the US what the US and Allies did to Japan and Germany.

The F-35 can win battles, easily, no argument from me there. But a war is different.

>Neither is taking down an F-35.

Maybe someone said that once upon a time about the B52 Stratofortresses until the Vietnamese shot down more than a dozen of them. Powerful war machines have been constantly brought low by hordes of weak and low tech enemies throughout history. I contend that is still the same now.

A true feat of manufacturing and engineering unmatched by any other military.

Also unmatched in cost and rarity. It has been proven many time that quantity is a quality difficult to beat. The US military has been gearing and shifting toward expensive and hard to mass produce war machines. We will see how well a million dollar tank fare against 100k dollars worth of drones. Or what happen when a 10K rocket need a CWIS to take down that cost 50K to shoot.

My best guess is that the hole was there before the mushroom grew. It is hard to find a biological species that need so much iron in their metabolism to the point it evolved something that can dissolve pure metal in a matter of 1-2 months. Irons are often used in enzymes and usually the ratio is around a few iron atoms bounded in a structure of thousands or tens of thousands of carbons and hydrogen atoms. To use so much iron implied that mushroom is literally chock full of special enzymes and I doubt it actually is.

More likely, something broke a hole there and the mushroom grew on it and capitalized on the broken edges to scavenge some extra irons but it did not punch that hole by itself.

We have been learning more and more about the importance of the gut microbiome, from emotions and intelligence to immune responses and hormones. It is my opinion that children should be allowed to explore and "get dirty" so to speak. It would train and expose their body to different elements and pathogens that build a balance and "natural" microflora. By babying and obsessed with keeping kids clean and spotless all the time, we may actually hurt them more in the long run.

A bit of dirt, some sneezes, a scratched knee, a few bruises, all of them are normal and expected. Kids should be allowed to have those, otherwise we might deprive them of a healthy life later on.

Anecdotally, I grew up in a not so clean place. Thinking back, I am surprised it was so normal back then with how unhygienic everything was by Western standards. You would think people living in that kind of environment would be sick and disease ridden all the time. Yet it was the opposite. Allergies were practically unheard of and stuff like a cold or flu were almost never severe enough to send anyone to the hospital. Basically, people I knew back then were a lot more robust and had less health concerns than most of my coworkers and neighbors nowadays despite a world of differences in hygiene standards.

IIRC, germ free rodents grown to be absolutely sterile for specific medical tests are extremely sick all the time and are so weak you would not believe they are genetically the same with those rats thriving in the sewer. I guess it is for the same reason.

A good introduction but it is too narrow on the "science" side and not enough on the regulatory side.

I can say this with pretty strong confidence: had the FDA not being so strict about its trials, lots of these experimental type of antibodies would have already received massived research and commercialization. But it is simply so much harder and so much more expensive to push something so different through the pipeline. A million things can go wrong, and even the biggest biopharmas can't handle that level of risk. Drug development, especially biologics, is so expensive it becomes incredibly conservative and resistant to changes.

Five guys in a garage can make a tech company that shakes the world. But 10 guys in a lab worth one million dollars trying to make a biologic drug will likely declare bankruptcy by the end of the year, at best. I would say by the end of the month is more likely.