If you are considered to be rich in terms of ways and means and generous of spirit as well, you really have it made in the shade. Traditionally speaking.
No one has to say it and plan it out loud. But the data will be sitting there. The incentive to improve the model for enterprise use will only get stronger. It doesn't take much for one engineer or team to go rogue to hack a benchmark. There was a whole cheating controversy with llama 4.
What does cheating benchmarks have to do with breaching financial user agreements?
To use your argument: all it takes is one whistleblower to get the company sued for billions of dollars.
What does unethical behavior have to do with unethical behavior? What does looking at data you're not supposed to look at have to do with looking at data you're not supposed look at? Please don't be obtuse.
People whistleblow on Meta all the time. Most recently they got fined half a billion for suppressing child safety research. Kids getting groomed. It doesn't make a difference to a company that has a net profit of $60 billion a year.
If you ever actually see the work or something charities on the ground and the people they help you might change your mind very quickly. They certainly do more good than another million in some index fund.
No that’s never happened to India you’re right. If you “support” relocation of a Ukrainian like Iryna Zarutska they might end up moving to Charlotte, NC where they get stabbed on a train. Road to hell paved with good intentions and all that.
No, but the UK did support the relocation of my great-grandfather who fought alongside the British against the Soviets and Nazi's leaving him no home to return to once the war was over.
He dedicated his life to the defence of his home, then to Britain and then spent 37 years in the coal-mines powering the country. All the while being a refugee.
I totally understand your worry about globalism, and to be clear: very human feelings about demographic shifts and change are being stoked by both sides of an ever polarising political spectrum.
But, I think it's important to remember that the reason refugee protections even exist as a concept is because there are real people who suffer, and we offer protection to those who need it not because we think ourselves superior: but because we hope that if the day were ever to come for us: that other nations would help us live our life so we could return home.
Unfortunately for my great-grandfather, even though he settled in the UK; he would have returned, but his part of Poland had become part of Ukraine, and Ukraine was the Soviet Union: the entity he went to war against.
I can tell you there will never come a time when American or UK refugees will be able to look to Poland or Ukraine or Afghanistan or India for rescue. It is that level of idealism and high trust naïveté that makes first world civilizations nice places to live but also dooms them to be taken advantage of by tribal mooches if there is no careful pragmatism.
Idk to me this is just redescribing what deep neural networks do without actually explaining why anything happens. I guess it "unifies" things but I am kinda over most unifying theories. Everything is Bayesian, everything is a graph or a group or some other fancy geometric structure, everything is a category. Ultimately the best framework is whatever is useful enough to explain what's happening in such a way that a practitioner can manipulate the model towards a desired outcome. In other words, where is the knob? The tool they share may be interesting and I hope to play with it to see what happens at different levels of noise applied to the labels.
Lastly, we derive an exact population-risk objective from a single training run with no validation data, for any architecture, loss, or optimizer, and prove that it measures precisely the noise in the signal channel. This objective reduces in practice to an SNR preconditioner on top of Adam, adding one state vector at no extra cost; it accelerates grokking by 5x, suppresses memorization in PINNs and implicit neural representations, and improves DPO fine-tuning under noisy preferences while staying 3x closer to the reference policy. [1]
I'm having a hard time understanding what IP of value could have possibly been generated in less than 8 months to make it worth $400 million. I guess it's just spending some stock into hiring some very smart experienced people to get ahead in the life sciences domain.
Yes, the US spends by far the most on healthcare per capita, and still gets the worst outcomes. It's not about throwing money at problems. You have to actually solve them (and want to solve them).
Instead of fixing the leaks in the plumbing we're just increasing flow rate at the top of the system. Right now the medical-industrial complex is a business that happens to provide medical services as a byproduct.
Fee for service encourages more medical intervention. It makes break-fix work more profitable than trying to address root causes. Each layer and sub-sector is for-profit. The supply of physicians is capped. Vertical integration can reduce the double margin problem. Specialists are paid more than general practitioners. Surgeries make more money than non-surgical treatment.
People expect someone else (their employer, the government, the rich) to eat the cost of healthcare. Third party payment makes patients price-insensitive. People unironically expect a cure for death. Drugs and medical devices are expected to be perfectly safe. Americans lead unhealthy lifestyles. Soda is the default drink in cafeterias and at parties. Car culture and car only infrastructure promotes physical inactivity. Kids are not allowed to roam. When there was a push to promote healthier school meals there was immense pushback.
This article basically says “yeah social security is not solvent and we probably won’t have a solution any time soon, and if we do have a solution it will be bad, but just don’t worry about it bro.”
I came to a different conclusion. That you should expect a 25% haircut through a combination of benefit cuts, tax increases, and not keeping up with inflation. That's bad, but Social Security is not going to entirely implode.
That’s the recommendation in the conclusion, and probably the advice I’ll follow because I don’t really have a better idea. Need to strike a balance between optimism and pessimism.
Social Security is a separate tax in a separate fund (invested in T-bonds). If you split that out of the budget, the budget looks even worse. The boomers retiring is just going to be bonds maturing without the holder (social security) reinvesting the money.
And yet that GINI keeps rising and cost of living outpaces inflation over the past 40 years in housing, healthcare, and education. If you are bottom quintile, maybe bottom 2, you are poorer now than 40 years ago on average. Recent policy changes seek to drop the third quintile as well.
How are you at the mercy of a group with no money or power? Genuinely curious how you logic-ed your way into this tenuous and seemingly ludicrous position
They have money. It’s taken from the rest of us and given to them. The rules you follow in life will be based on the behavior of the bottom quintile, the taxes you pay are to support the bottom quintile, the greatest risks to your life and property will come from the bottom quintile, the dearth of comfortable public spaces is because you have to allow the bottom quintile to be there, our zoning laws are developed for fear of the bottom quintile.
The problem is the promotion of values and behaviors that plague a low-trust society. I think making excuses for it is truly inappropriate and immoral.
reply