I can confirm - I woke up to the resolution to my two hardest problems during PhD. Three, if you count "I should look for this kind of inequality" (which did turn out to exist), but I think that's more of an _idea_ than a solution.
The hard part is paying attention to it. With enough attention your mind will fix it.
Thanks for sharing your opinion. The first difficulty in real world (even idealized) politics is voicing a consistent independent stance and thinking through the implications. Many cannot even make it that far.
The second and greater difficulty is that realizing that a solution that is politically untenable is not a solution, it's a campaign slogan. I don't know how we get people to move past this difficulty.
I assume you are suggesting that a tax on the rich is not politically viable.
When the structural violence that permeates our society finally manifests in the only violence the lower class can execute - this non-viability might change.
It will be a harrowing time- and I hope we avoid this.
The billionaires will cease to exist one way or another.
I did this too! For months (almost a year) I used descriptions, pictures, and measurements of food to get rough calorie counts. My diet is pretty simple and repetitive.
I would occasionally check the estimates, maybe once every few days for meals I wasn't already pretty sure of, and it was generally accurate. Where it was extremely inaccurate was on portions, and anyone who has dealt with computer vision could tell you, you can't get scale from a picture. So I'd have to weigh some meals or ingredients, which would generally make things more accurate again.
So, I think it's possible, but you need multimodal data and grounded with regular checks.
About once a week I ask ChatGPT to give me a reasonable diet for recomp with weight loss. It consistently insisted I have at least 7 meals consisting of at least 30g of protein per meal, but the protein source can't be whey or casein. When I ask "why" it cites a bunch of studies ... but most of those "studies" are N=1 of a college or Olympic level athlete. If, instead, I grab a large scale lateral analysis, it says "3 meals" with about 1/2 of the protein.
It'll defend both sides (mutually contradictory) to the death. NOTHING will budge it from its initial stance.
To be fair this is a reflection of the general state of nutritional science and the actual answer seems to be "it depends on your genes".
Some people do well on 6 small meals, others do well on no breakfast and two large ones. Studies can't tell you anything useful about that, you have to experiment and find out what works best for you
The answer does not really depend on genes. There are personal preferences, there are sex differences (women prefer more carbs), and the biggest component is where you are and in which direction do you want to go to.
But in terms of physiology the answer is quite clear:
1. The protein is the most important macro to get, no matter if bulking or cutting. It is the building block.
2. Whatever the amounts (0.8g-1.8g/kg of bodyweight, depends a bit on a situation and the willingness to lose some potential marginal gains), try to divide your daily protein somewhat evenly between meals.
3. Pareto principle, you get the most benefits by having 3 meals. 4 if you really care about small differences and want to optimize. 5 meals give negligible additional benefits, for professional athletes who want to be anal.
4. So basically eat at least 3 meals and up to whatever works for you practically speaking.
Optimum meal timing in particular I believe is heavily influenced by genes - I have friends who never eat breakfast, survive on black coffee until 1PM, then eat a lot in the evening and feel good doing it. If I do that I feel terrible.
So yes eat 2g/kg protein but the best way to time that in terms of meals, best specific foods to eat etc is definitely influenced by your genes
But physiologically speaking there is no difference in the optimal recommendation in terms of protein synthesis: protein should be evenly divided between evenly spaced meals.
If people decide that this non-optimal way of life fits them better, then go for it. But let's stop with the "I feel better this way, therefore biologically my body works differently than any other human body." BS.
I think the analogy (and it is not to be taken literally) is that of "commoditized processes".
Nowadays we don't build bridges to suit the site, we choose sites to accommodate bridges that we basically build identically via a few designs.
Connecting back to s/w AI can do the standard stuff ok as long as you test around the outside of it, so you might want to hone your judgment about how you build systems so it uses the stuff AI can do well, vs "building for the site". The gains are productivity. The losses are efficiency (the problem must go through some extra steps to meet the process where it works). Same as any engineering problem at scale.
Yes I believe GP was focused on the catastrophe part. It's very likely correct that our CO2 emissions are warming the atmosphere ocean etc, but it's not clear that runaway warming is inevitable or that life or geology have feedback mechanisms that turn an exponential into an S curve. That is, after all, basically what natural selection tends to do. Turning the table again, even if there are corrective factors humans might have immense suffering before it stabilizes. So we don't know.
You didn't ask, but my opinion on it is that we'll probably stabilize on a cleaner energy source and find natural countermeasures when suffering ticks up. Any top down pressure to change things whole cloth seems doomed, no matter how benevolent. We're closed loop creatures.
If you think a 2 year old is doing deep learning, you're probably wrong.
But if you think natural selection was providing end to end loss optimization, you might be closer to right. An _awful lot_ of our brain structure and connectivity is born, vs learned, and that goes for Mice and Men.
Why not both? A pre-trained LLM has an awful lot of structure, and during SFT, we're still doing deep learning to teach it further. Innate structure doesn't preclude deep learning at all.
There's an entire line of work that goes "brain is trying to approximate backprop with local rules, poorly", with some interesting findings to back it.
Now, it seems unlikely that the brain has a single neat "loss function" that could account for all of learning behaviors across it. But that doesn't preclude deep learning either. If the brain's "loss" is an interplay of many local and global objectives of varying complexity, it can be still a deep learning system at its core. Still doing a form of gradient descent, with non-backpropagation credit assignment and all. Just not the kind of deep learning system any sane engineer would design.
I don't know what you mean by end to end loss optimization in particular, but if you mean something that involves global propagation of errors e.g. backpropagation you are dead wrong.
Predictive coding is more biologically plausible because it uses local information from neighbouring neurons only.
There are highly politicized blogs which can discuss this further and offer opinions as to the correlation.
When DJT talked about cancelling the Dept of Edu, I got ahold of all the teachers I knew and asked them what impact it would have, and their response was mostly that they would lose their school lunch benefits.
> Funding per student is on the rise, or level on inflation-adjusted $
That's at the state level. But that doesn't account for the explosion in admin salaries and positions. The actual money a district spends on each student has been going down every year. Those funds are going more towards admin activities.
> I got ahold of all the teachers I knew and asked them what impact it would have, and their response was mostly that they would lose their school lunch benefits.
Teachers have a very poor understanding of where their funding comes from. Most just assume "property taxes", but it's far more complicated than that. The department of Ed provides a lot of funding to states that is passed through to the schools. They also enforce the education titles.
Cutting the department of Ed may not have a direct, immediate impact on classroom teachers, but it will have a large downstream effect in a few years.
Student/teacher ratios have gone down, not up over the last few decades. This isn't a lack of funding.
Teachers are put in an impossible position with students who come from homes where the parents don't do their proper jobs. It's never been easier to be a neglectful parent. Your child will be entertained non-stop by an iPad and a video game system. They won't get bored and bother you. You can send them to their room and do whatever you want if you don't care if they are sleeping or not, as long as they are quiet.
The "iPad babies" are an epidemic in schools.
Source:
My sister is a K-12 educator in a poor, rural public school system in southeastern Virginia.
In recent years, she's seen a surge in students who are sorted, improperly, into special education classes. These are students that exhibit symptoms of various learning disabilities, but these symptoms heavily overlap with the symptoms of children who are sleep deprived and over stimulated by dopamine activating content on the devices they are addicted to.
The single variable that actually matters when it comes to school - and nothing else matters until this one is fulfilled: The quality of the peers who make up the student body.
Or put another way: The quality and involvement of the average parent.
A school can absorb an extremely small minority of "problematic" students if the rest of the student body is stellar, but that's about it.
There is not a single thing any public education system can do to counteract that simple fact. If the average student in the classroom is uninterested at best and troublemaking at worst, it doesn't matter how good the teachers are or what the ratios are, or if the classrooms are old and busted or brand new.
Until society becomes serious again, this problem will only get worse as education continues to be a political and culture war football. The best realistic thing I can think of is take a look at nearly all other western social democracies who have much better outcomes and immediately implement student academic tracking. But that would be politically impossible to do in the current state of the US.
I fear that things are going to get far worse before they get better. You could 10x the primary school education budget and likely continue to see worsening results.
When I went from private (poor) primary and middle school, to a rich suburban high school, to a poor inner city high school back in the 90's this was self evident. I didn't think it could get much worse than that, but the administrative and political classes figured out how to wildly beat even my exceedingly low expectations.
If you ask boomers they'll be far more likely to tell you dad was out working 16 hours in the oil field / carpenter for the housing boom or something like that. Mom has no time for you either, she is busy with the 4th baby. Kid gets a nice belting for bad behavior and other than that, be back before the street lights come on for a dinner conversation and then left to your own devices before bed.
I think if anything parents are more involved now than they used to be.
The most obvious difference to me other than ipads/social media is we don't beat kids anymore and we give them way less autonomy.
Yes. When I was ten I had friends ranging from age eight to fourteen that I regularly hung out with, and older people in their forties or fifties that I would drop by when I saw them outside to learn things from. There was a hierarchy of responsibility in our group where the oldest kept track of the ones younger than them, and those kids kept track of the ones younger than them. Beyond that we had no supervision because everyone knew someone who disapproved of something and would tell their parents, whether that be my same age peers disapproving of cursing or the eldest kids disapproving of everyone going to someone's house uninvited. That risk of strain on the friend group kept everyone in line.
Nowadays parents are very strict about the age gaps between their kid's friends, especially with how older kids usually know how to get into risque stuff online. They aren't exposed to differences of opinion and ability as much in real life, and that somewhat hinders their development. There's nothing that can teach you patience like trying to calm down one of the younger kids so you don't get kicked out of a friend's house during the basketball game. Just like there's nothing that motivates you to get better than your six foot two friend intercepting every single pass to your receiver.
And this isn't even getting into the hobbies, interests, and skills kids can learn just by watching adult neighbours. While this year I'm seeing more people outside doing things, for a long while everyone was inside. That meant there weren't older people outside working on cars, tending to their gardens, preparing their boats for fishing season, or just sitting around talking about activities from the past that might be interesting. Kids are more likely to take an interest in a new activity if somebody they know does that activity, because that person is much easier to ask questions and directly show problems to. If they see something online it will probably be a momentary passing interest that they'll forget by the end of the day because once that video's gone so is their interaction with it.
The school system is going to crash out in the US. The public school teachers will readily share symptoms ("enrollment going down", "2X the IEPs of 3 years ago", "non-verbal ipad kids", "kids only sleeping 4 hours a night because of ipads", etc.). As everyone with means or time escapes, the system increasingly distills problems and legitimate special-needs cases while no longer spreading them out among cooperative kids, and teachers will continue to burn out in such a thankless environment.
At varying times in various places, public school is or will become just like riding the bus: technically a viable option for a needed service, if you have no other choice and are ready to suffer in a place that tolerates all manner of dysfunction.
Yes to this! So many people turn a blind eye to the critical role parents play in supporting teachers holding kids accountable. And I get it, holding kids accountable is very, very challenging, but that's the gig people sign up for when they decide to have a family.
And I'm a former high school teacher and my wife is a current high school teacher so I've experienced all of this first-hand.
It's also becoming increasingly more likely to enter into college with lower relative and absolute high school performance.
Perhaps some wonder why they should try so hard in HS, when most anyone that graduates can get into college, and no employer is asking a college grad what their high school grades and scores were.
There was a time back in the 60s or 70s or earlier when anyone that graduated HS could get a decent job. And a time now where most anyone who wants a decent job, must complete college or trade school. The latter are increasingly becoming less correlated with HS performance. The importance of HS performance needed to succeed is regressing back towards what was needed back in the 70s or before, so long as you actually graduate so you can go on to further schooling. In the 80s -00's was a time where you where the ladder was shut off if you didn't go to college, but going to college was far more correlated with having the highest marks.
Students should "try harder" in high school because the point of school isn't to "get into college". The point is to learn how to learn and become better at problem solving. It is my opinion that being a good problem solver is the entire point of education.
In an era of declining birth rates and thus fewer students graduating from high school, of course the third-tier private colleges are going to lower their admission standards in order to survive. In the long run this won't work because employers will eventually figure out that degrees from those colleges are worthless. But they'll keep up their grifting for a while, and leave a lot of mediocre students stuck with huge debts they can't pay off.
For a long time, college education was the easiest way to legally discriminate against applicants. The signal is weakening and the expense of exhibiting the signal has skyrocketed.
> That's at the state level. But that doesn't account for the explosion in admin salaries and positions. The actual money a district spends on each student has been going down every year. Those funds are going more towards admin activities.
Ok? Seems like that's more of a problem than the funding. Or whatever is causing that is more of a problem, but it does a disservice to the general argument of "kids aren't receiving the same level of care" argument to blame a drop in funding--especially when it was so easily falsified.
It's crazy to pay income taxes to the federal government and then have the Department of Education turn around and grant that money back to the several states so they can use it to fund public school districts. A lot of those tax dollars get wasted along the way. Better to cut out the middlemen and send property and/or income taxes directly to local governments, with some state level aid for poor areas with low tax revenue.
That leads to a problem (which you partly addressed with the state level aid) of linking education funding with the wealth of the area which I suspect should be inversely linked i.e. poor areas need more funding and wealthy areas require less as the kids are typically in a more stable situation and aren't skipping meals.
> At the same time, school scores started to sag after 2014
That's around the time a bunch of districts in a state I lived in at the time had multi-year teacher pay scale freezes due to budget crunches. Not saying it's necessarily connected to the scores dropping, but still.
Total spending across the country may be high, but it's very much state-by-state and local how much is spent and where it goes. Some states pay teachers pretty well. Some states, the pay really is pretty awful. Some states are OK on staffing levels. Others are in an ongoing staffing catastrophe that's forcing them to cut school days to try to get by.
Meanwhile, school performance is heavily tied to home life and broader community support for students' families. That's why all this effort to improve schools hasn't been as effective as one might hope: the attention needs to go toward much harder problems that have little to do with schools and are really hard to get any progress on in the US. Worker protections, better and less-stressful "safety nets", better policing and a better justice system. That kind of thing. I'd look at least as much at what's been going on with those, and with security and home life for those in the lowest three quintiles of household income, as at schools themselves, to try to find reasons for trends like this.
Because the internet addictions/phones everywhere mean the average dumbass kid is reading more actual text than any kid beforehand on average. This is why the missippi miracle is happening. Well, that, and reducing the amount of actual corporal punishment administrators can doll out (paddling students in public school is legal in the shithole Deep South )
I met a forensic accountant recently who mentioned a corruption investigation she participated in involving a school district nearby, several high-ranking board members and admins were on the take. She pointed out the futility of the project, it was a large sum of money for a school district, but nothing like your headline-grabbing Medicare scams. She wound up leaving the investigation due to threats to her safety and took another job. It felt like one of those unresolved endings to "The Wire".
I have had enough insight into enough school districts that I'm confident lots of them are hotbeds of corruption. Mostly at the upper admin level (superintendents and such). Kickbacks for contracts, hiring absurd numbers of assistants and secretaries to the point that one wonders what work remains for the top dogs, creating do-nothing decently-paid positions for people they're having affairs with. That kind of thing.
It doesn't even have to be outright corruption or fraud. It can just be "fraud lite" as I have come to call it. Won't actually qualified as fraud in any "academic study by the experts" but anyone who stuck their nose in and witnessed the ongoings would immediately call it for what it is.
Could simply be doing the bullshit "spend down the year's IT budget on stuff likely to sit in shipping crates at district HQs until it gets e-wasted". The latter being one of the few I directly witnessed - millions of dollars of Cisco gear sitting there for 5 years before it was trashed. Never needed in the first place. I have no reason to believe anyone was on the "take" for it - just general incompetence and grifting to keep one's Very Important job going for internal politics.
This was for a district where a few million could easily have paid to fund a district-wide music program that was recently cut, among myriad of other in-the-classroom things.
The older I get and the more I witness things like this, the more I understand why a large and growing segment of society has completely tuned out the "experts" trotting out studies and reports. Those have largely been weaponized, and the erosion in trust of both institutions and expert knowledge may now be terminal due to it. You can only be told the sky isn't blue by so many experts until you tune them out entirely.
There hasn't been enough said about the corruption of public life in the US. (And elsewhere.)
It used to be this kind of thing was - maybe not exceptional, but certainly not expected.
Now it's common but underreported.
So there's a kind of dream world where "education" and "health" are still considered official public goals. But the reality is that government procurement is mostly grift and corruption. There's been an epic collapse of almost any kind of public service ethic in favour of opportunism and profiteering, sometimes covered over with religious/moral pretexts.
I've just been assuming it's all gotten way, way worse over the last 20 years or so, too. One of the main things keeping it even slightly in check was local newspapers and TV stations with actual reporters.
Those are all gone, either shuttered or snapped up by huge companies that fired most of the staff and are milking them for the last money they can provide, or using them to distribute propaganda (e.g. Sinclair), and nobody's ever going to (be able to) do a proper accounting of how much the resulting waste and corrosion of public trust has cut into the actual overall cost/benefit of this whole "Internet" thing.
I remember hearing David Simon, creator of The Wire, predicting this (fall of local news enabling unchecked corruption). Here's an article on it from nearly 20 years ago:
> "Oh, to be a state or local official in America over the next 10 to 15 years, before somebody figures out the business model," says Simon, a former crime reporter for the Baltimore Sun. "To gambol freely across the wastelands of an American city, as a local politician! It's got to be one of the great dreams in the history of American corruption."
One of the seasons of The Wire is largely about a major newsroom (the Baltimore Sun, unsurprisingly) taking its first hard punch from the collapse of the news market and unchecked M&A activity, so I'm not surprised he commented on it elsewhere too. God, what a great show.
I'm not sure there is a viable business model for local investigative reporting waiting to be discovered, any more. At least not in the US, not in mid-sized or smaller markets. It's semi-functional in rich, dense cities. Might remain so for a while longer. It's just everywhere else that now has no watchdogs aside from the occasional, lazy, probably partisan look-see from state regulatory agencies, and maybe resource- and access-starved hobbyists if they're lucky. The pros are gone. A few still watching big national-scale stuff (bigger audience!) but all the smaller parts of the system have gone dark.
Corruption is not new- in fact looking at US history it appeared to be the norm. Tammany Hall, railroad barons, the Prohibition, Standard Oil. There were just a brief few decades after WW2 when it slipped into the background.
It happens in the private sector too. I was involved in procurement at a megacorp for several years.
At one point one of my colleagues asked for assistance in getting an order of 500 iphones approved. As "spares".
Fortunately the corp had a policy that phone purchases needed to have a named individual declared.
I declined politely to assist.
It was common to see certain mid level execs churning through 2x - 5x the equipment of IC's (who would never get out-of-lifecycle approvals anyeay) and some quid pro quo stuff. As a fraction of their total comp it was modest ultimately, and for this reason my boss advised me to keep my mouth shut.
There’s probably a feedback loop: as people have become convinced that the government is only useful for corruption, that becomes an expected perk of the job.
Unfortunately, I don’t see a way out of that loop. Move to a state that still has some civic pride I guess.
The way you get out of that loop is by creating immense pressure from the outside until the governing system breaks, then supervising the reconstruction as an outside power until it can function by itself again. The issue is that there's a very high risk of it suffering malformed development during that reconstruction, or even worse it's abandoned early and never even builds the functionality needed to sustain itself. The risk is so high that people prefer to let the system degrade with the hope that it will eventually halt or in the slimmest chance even regress to a better previous state. Meanwhile the success rate is so low that I can think of a myriad of failures off the top of my head including Panama, the Kingdom Of Italy, Albania, the American South during Reconstruction, and Indonesia with the only success coming to mind being Japan.
That explosion plot is pretty bad. No y axis. Unclear if it includes tuition and loans (which are paid or owed by students). Loans are ~50% of "ED appropriations", and only ~$21 million was distributed to students in 2021. But in the plot is looks like spending was around 150 billion (hard to say with no y axis) but ~50% was loans? And the source is just a vague Dept of Ed, with a link at the very bottom of the page to every single table published by the Dept of Ed, so have fun checking the source.
I'm not criticizing your take, although I suspect teachers might lose more than their lunch, just pointing out how terrible the plot is.
You should not adjust for inflation or even for wages, but for cost of employment. The way health insurance works in the US makes public sector jobs with average wages and good benefits expensive to the employer.
> When DJT talked about cancelling the Dept of Edu, I got ahold of all the teachers I knew and asked them what impact it would have, and their response was mostly that they would lose their school lunch benefits.
The hard part is paying attention to it. With enough attention your mind will fix it.
reply