Very interesting and well-explained. Given that the research has been out for two years, any interested data collectors have considered this! Forget hackers, this an exploit for enterprises and governments!
Could websites concerned with privacy deploy a package that triggers interrupts randomly? Could a browser extension do it for every site?
Websites doing this would have to be careful about it: they might become the only website triggering a lot of interrupts randomly, which then makes them easy to identify.
Our countermeasure which triggers interrupts randomly is implemented as a browser extension, the source code for which is available here: https://github.com/jackcook/bigger-fish
I'm not sure I would recommend it for daily use though, I think our tests showed it slowed page load times down by about 10%.
I'm on safari/macOS, and many of the counting related demonstrations did not vary as much as claimed -- some did, with significant computer use, but I'd bet some mitigations have been implemented already in Safari.
We spent a whole gray day at Bletchley and TNMOC and I appreciated the connection between Bletchley and TNMOC given the context of WWII. To me Bletchley is more about process, a metaphorical "Scrum room" of one of the most important math and science programs in history. And taken together, they encapsulate sort of a catalyzing moment that computing prehistory transitioned to computing history.
My history-teacher wife liked Bletchley a lot more than TNMOC, where I lingered too long, but I did like both. Even though the Bombe replica was down for repairs that day.
Definitely if your interest is World War II, or the Intelligence process, the human side, any of that - the Bletchley Park exhibits are key and TNMOC is at most a side visit, whereas if your interest is computing, TNMOC is key and there's no reason to do more than pencil in the rest of the Bletchley Park site as a possible extra.
I don't know about Science. There's a lot of math and logic puzzle solving at Bletchley, but the other crash projects of WWII have a lot more science, Los Alamos obviously, translating from "In principle nuclear fission is a more powerful bomb than anything previously made" to an actual weapon you can use to destroy a Japanese city. But also the invention of the Cavity Magnetron - a little box can make enough radio waves to make a radar for your night fighter so you can figure out where the enemy planes are relative to you - or it can use those waves to heat a delicious baked potato in a few minutes...
it does seem like working on the Bombe got Alan Turing really deep into thinking about hardware. he'd previously played around with building some logic circuits but seems like he didn't go deep into it before Bletchley, before that he was mostly a very pure mathematician.
Turing was interested in lots of things. As a coincidence one of the small jobs I did many years ago was to develop the Turing Archive's web site, which I'm glad to see has since been substantially renovated: https://turingarchive.kings.cam.ac.uk/ - For that work we had a lot of high resolution images of non-Computing stuff Turing cared about, including Morphogensis (basically, why things are the shapes they are, for example why is a rose petal shaped that way? Why stripes on giraffes but spots on dalmations?).
So mathematics yes, Turing was always interested in that, but always applicability was on Turing's mind. In the early twentieth century Turing's "machine" in a paper he wrote at Princeton (before the war) was just an idea, but er, obviously with the exception of the need for an "infinite" paper tape you can realise Turing's machine, it's a computer, familiar to everybody today.
Even after Bletchley Turing wasn't a Software Engineer. The meta-applicability isn't something which would occur to a mathematician. That took Grace Hopper. Grace understood that the problem her people were tasked with (mechanically convert instructions to the code for the actual machine) is exactly the same sort of task the machine is doing anyway, and she invents the Compiler.
The purpose of the article is persuasive but the HN title is ambiguous and reads much more expository. I’m a New Yorker who walks, bikes, and drives, in roughly that order and it was clear to me that the author is pro-bollard.
I work in news technology, including with many local news organizations, both corporate and independent, including in California. Google does not support all news organizations equally, and this seems designed to gain some leverage over organizations that do get a lot from Google. Their Google News Initiative is on its surface just a training platform for publishers to learn how to use Google’s tools but they’ve done quite a bit more for some publishers. This feels like an attempt to gain the vocal support of publishers who have been blessed by Google’s beneficence, many of which are earnest non-profit organizations who might take the bait about big bad hedge funds. I don’t know how they select who to help and who to ignore and our attempts to engage with them on behalf of publishers have had mixed results.
But publishers’ collective frustration with Google is quite high. Given the implications that “it would be a shame if something happened to your nice journalism website” coupled with the appeal against the big bad hedge funds and ghost papers, it’s sort of a clever position but I’m not sure it will work.
As far as I'm aware, the state of play in tech is far more welcoming to formerly incarcerated people than this thread would imply. Justice Through Code at Columbia University is designed to place formerly incarcerated and criminal legal system-involved individuals in tech roles, and has had a lot of success placing its alumni at big tech companies [1]. And this is largely for entry-level tech workers.
Checkr is a commonly used background-check tool, especially in tech, that allows for those with criminal histories to provide context for what's on their record [2], I'm curious if you've encountered it specifically.
To your post about being ghosted, that seems unfortunately to be a common theme in this period of staff contraction that may not be limited to those with criminal records [3], but reneged offers is a bummer, I'm sorry that's happened to you. There's a theme of "owning the narrative" among some formerly incarcerated people that may be worth considering.
Last, a useful resource on humanizing language for those of us without criminal justice histories [5].
Totally. During the early months of the pandemic when NYC was hit super hard, Samaritan's Purse set up a tent hospital in Central Park directly opposite the main campus of Mt. Sinai which was a marvel to see, I would run and cycle past it nearly every day. Another of my common running routes took me past the USNS Comfort, docked at Pier 90, and the Javits Center, which hosted yet another modular hospital. Then a little later, a modular morgue on Randall's Island, which is now the site of a modular migrant encampment.
I later read that because of strict policies relating to a notion that only the "simple" cases, of which there were few, could be transferred, the temporary hospital facilities ended up underutilized. But the ability to surge capacity in a specific area is hardly constrained to parts of the world with fewer hospital beds. Public health crises and natural disasters in the USA are perfect candidates for this technology.
I hope we can test this quickly, and if it works, buy a bulk pack.
I feel like in the parent comment coup is sort of shorthand for the painful but necessary work of building consensus that it is time for new leadership. Necessary is in the eye of the beholder. These certainly can be petty when they are bald-faced power grabs, but they equally can be noble if the leader is a despot or a criminal. I would also not call Sam Altman's ouster a coup even if the board were manipulated into ousting him, he was removed by exactly the people who are allowed to remove him. Coups are necessarily extrajudicial.
It also looks like Sam Altman was busy creating another AI company, along his creepy WorldCoin venture, wasteful crypto/bitcoin support and no less creepy stories of abuse coming from his younger sister.
Work or transfer of intellectual property or good name into another venture, while not disclosing it with OpenAI is a clear breach of contract.
He is clearly instrumental in attracting investors, talent, partners and commercialization of technology developed by Google Brain and pushed further by Hinton students and the team of OpenAI. But he was just present in the room where the veil of ignorance was pushed forward. He is replaceable and another leader, less creepy and with fewer conflicts of interest may do a better job.
It it no surprise that OpenAI board had attempted to eject him. I hope that this attempt will be a success.
No. The sunk cost fallacy involves sticking with something you know is the worse option (sticking with that option is the worse option) by appealing to the investment already made in that option as a justification. It does not mean that all appeals to investment are fallacious. If I invested 5 years in graduate school, and I'm a few months away from receiving my doctorate, it would not be categorically fallacious to argue that since I've already invested this much time, I should stick it out even if I'm thinking of pivoting. You would be losing a degree for what could be a vanishingly small benefit of leaving.
In line with that, the comment above is asking for some reason that could explain why someone who has invested so much time in something would leave so abruptly. A person who has worked on a project for so long derives (and provides) many benefits for having been involved for so long, and even when people change course, they don't typically do so this abruptly without a proportional reason. Hence, the mystery.
I have a lot of Blackberry nostalgia; I was 23 in NYC, 2007ish, and my work-issued BB Curve (later a Bold) plus my personal original iPhone were a combination that I was so happy with. The Blackberry could tether to my early MacBook pro in the remotest of locations, and I could easily SSH into the (real) servers I ran. Never used BBM but played Bricks for hours and hours. I wasn’t nearly as satisfied with an iPhone-only solution until 2017 or so.
I want to endorse the process and maybe not worry so much about the result. I started on a similar journey in the pandemic of making pizza for my wife and kids every Friday night. They all receive their own (small) personal pizzas, which I knead, roll, and transfer from a pizza peel to a pizza stone. Starting from a NYTimes recipe for Roberta's pizza dough — a pre-children favorite of ours — and iterating from there, I became confident over dozens of repetitions. My pizzas now are very crisp on the bottom without being burned on top, and I can "sense" when the pizza is done, no timers needed. I can feel when the dough is hydrated, I can see when it's rolled flat.
All this isn't about knowledge that can be imparted in a book; it's frankly about kaizen. The art of doing it a little better this time than you did it before. If you do it 2% better each time, after ~35 times, you're twice as good. Don't get me wrong, I love (love!) the open knowledge here. But OP cannot put in the reps for you. Only you can put in the reps.
I love this. Also, in my experience doing the same thing with recipes from pancakes to roast chicken, I think it’s not just about kaizen—it’s also about the “tacit knowledge” that can’t necessarily be communicated through words.
I definitely wasn’t doing it 2% better each time — I was experimenting, trying out new things, seeing how they changed the outcomes, and building an intuition for how stiff the pancake batter was, what color it was, how much the chicken skin glistens, etc. When I tried something new, sometimes it was 2% better, but more often it was 25% better or 40% worse. Either way a success, because I learned what kinds of things were likely to work and what weren’t.
The “reps” help you not just get better, but (as you describe!) they build your mental connections between what you see, smell, and feel, and results. You start to recognize when things look or feel a little different, and adapt.
Honestly it’s a lot like developing expertise in programming!
This is where an experienced cook in the kitchen with you now and then can be so helpful. A book might give some pointers, but they can say "this happened because of that".
If you don't have that, try to not vary more than one thing at a time.
The problem with cooking books is that they represent the author's experience, local products and equipment. Different flours and yeast behave differently and ovens are not the same making a recipe just a starting point.
Could websites concerned with privacy deploy a package that triggers interrupts randomly? Could a browser extension do it for every site?