HN user
https://www.nature.com/articles/s41550-025-02564-0
The paper has more details. What's interesting to me is that the key innovation isn't the deformable mirror but rather the design of a wavefront sensor that focuses on coronal features (instead of the "grain" on the solar surface prior systems used).
But somewhere along the lines we're going to have think about the R&D of the big picture in non-hand-wavey ways.
How?
The article feels like it was written by ChatGPT, and it doesn't explain anything about the "special reflectors" or the overall theory behind the optical design that enables shadowless lighting.
https://infiniummedical.com/key-factors-for-clear-light-in-t...
This page has better information
Outcome-oriented design involves orchestrating experience design with a greater focus on user goals and final outcomes, while strategically automating aspects of interaction and interface design.
Dark patterns is another type of "outcome oriented design" where the outcome is the user subscribing/not canceling/paying needlessly. Adding "strategic automation" will doubtlessly just make this problem worse.
Mine has proven equally reliable to a mechanical toilet, while providing considerably more comfort. This actually makes sense when you consider all the mechanical complexity in a typical toilet is getting replaced with a series of solenoids. Flush is now just energizing an electromagnet, rather than the gravity-powered dance of floats and valves you're used to.
There's also fun to be had trolling toddlers. Sometimes when my 2yo closes the toilet I'll sneakily open it with the remote and she'll yell at it "NO!" and slam it shut again.
There was a good Chinese Pod episode on this but I can't find it now. One of the things I recall was to express emphasis, make the tones really obvious (almost like a beginner might, consciously speaking them).
If your operation has reached the point where milliseconds matter..something likely has gone wrong.
I'd be curious to see data on how other cognitive metrics change in absence of atmospheric nitrogen. Abstract reasoning / mental math abilities / etc. It's possible some actually improve, just like I might rush and make more errors while doing something after drinking a few coffees.
It'll be hard to make traditional motor windings out of this particular material because AFAIK it's a ceramic, but perhaps with thin films on flex PCBs it would be possible.
I'm imagining a future where a superconducting layer on a PCB is just another checkbox you can choose when ordering small runs of boards.
[ ] 1 oz copper
[ ] 2 oz copper (+$2)
[X] 10 micron LK-99 (+$10)
Another thought - I think the first place we'll see this widely rolled out is in IC's (waiting for the Asianometry video on it). IC's are already planar, they're small so exotic materials aren't a big contributor to costs, and they're very power dense. Replacing a metal layer with a superconducting one could enable greater gate density and potentially significant improvements in efficiency. I don't know by how much because switching losses are probably where most energy is dissipated, but it's an incremental change that seems compatible with the process.
They've optimized the fitness function "cuteness^2 - propensity to bite". So even when they're tearing up my arm it's still so adorable.
The paper glosses over the technical design of the solar shield spacecraft. I'd like to understand how they will maintain their attitude given how huge they are. Would gyros be enough to counteract radiation pressure?
Another perspective if we consider AI akin to nuclear weapons (a simile many have made already): If the law-abiding cease development and malicious actors do not, the worst case may be that the resulting AI is non-aligned with humanity's interests. If the "good guys" develop an AI with alignment as an intention, then there are at least good AI to counter the bad AI.
Of course they may all ultimately end up colluding to destroy us. My point is it's not just as simple as "someone who's behind might catch up and make more money".
Key point I think should be more prominent:
The effect was on cultured brain cells. The reason brain tumors and other disorders of the brain are so hard to treat is that the blood-brain barrier keeps many big molecules out. That doesn't mean the molecule isn't therapeutically relevant, only that there's a ways to go to making it useful for actual humans/mice and not just cells in a petri dish.
Still a fascinating discovery and I look forward to more research.
Edit: Not totally true, the actual study link has more details.
We further found that dietary supplementation with H. erinaceus crude extract significantly enhanced recognition memory.
The human BBB may still be an obstacle but I'm sure we'll see more studies soon.
That's sad. My dad had a subscription to these, so I read them quite a bit when I was younger. I think I still have a few lying around.
At the same time, there's more exciting stuff than ever happening in the astronomy world these days. Plug for Anton Petrov (https://www.youtube.com/@whatdamath) for those who like consuming news in video format.
... reaches up and flips the pressurization switch, cutting off bleed air to the cabin. The airplane rapidly begins to depressurize ...
I cannot believe that a "kill everyone" switch actually exists, and if it really does this seems like a bug. Especially because not hitting the switch would also kill everyone.
My pipes are all copper, and the hot water contains a non trivial amount of lead (from the solder used to join the pipes). I had it tested - around 20 ppb.
We don't drink or cook with the hot water, but in theory if the plumbing was PEX we could.
The article tries extensively to dance around the fact that all else being equal, if your head is hitting the ground you're better off wearing a helmet. In the worst cases, you aren't worse off. Acceleration is what kills, and because you can't change the starting speed nor the end speed (0 m/s), all you can change is time.
That doesn't mean they have to be enforced legally, of course. So rather than attacking the merits of helmets, the argument made in the article would be more compelling if it focused on the unalienable right of humans to get severely injured doing stupid stuff for fun or profit.
A few years ago that would be true. These days, minimum wage workers are getting harder to find.
But that's only if the search is constrained to immediate locality. The "great" thing about this concept is that we can hire minimum wage workers overseas to control robots in America (latency be damned).
I'm skeptical there's any value being generated here then. Who are the customers? How do they know this fancy looking biological design is any good at all?
It's fine not to actually test fire it, but in that case prove it with simulations. Demonstrate that it's superior to existing designs, or at the very least that it's functional.
I think adding another animal to the situation would only exacerbate the problem. Let's say we add lions. Then not only goats are eaten, but other animals.
Hunting goats to local extinction is the cleanest and most humane way to deal with them.
We'll fix it till it's broke!
Exactly - and that maximum is the "minimum maximum" across all batches of the device. You might get a particularly resilient one that handles the 10V just fine.
The encryption idea makes way more sense, and the means to short out the cap doesn't even require the device to be powered. The only implementation challenge is that now instead of combining a simple circuit with an off-the-shelf mass storage controller, you need to either find one that can read a key from an external source or roll your own using a micro (maybe this is easy with existing libraries?). But if I was a journalist concerned about such things I'd rather pay a premium for the right solution than worry that the police might be sweating from the heat or that the chip in my drive happened to be especially resilient to overvoltage.
Many comments are permutations of different suggestions do here's mine:
Get the business to buy into "fixing" this before doing anything. Convince them to hire more, sounds like the current team is already swamped.
If you don't get business buy in, it may be the wrong place for you.
The best approach is to assume there's a renegade employee constantly trying to screw the company over. Granularity of permissions should be set to minimize the blast radius to the absolute minimum they need to do their job.
The powershell script is a minor part of the screw up. The real issues are multitude...
1) hardcoding actual production credentials in a script at all. Seriously what the fuck.
2) Thycotic not enforcing MFA for the keys to the kingdom admin account. Even my cellphone provider has better security.
The root cause is likely the assumption that the VPN is sacred. This needs to die asap - your internal network should assume the VPN is wide open and secure everything accordingly. Defense in depth not eggshells.
That profile in particular is interesting. The first two tweets appear to be human originated, and date from December 2021. There are no tweets from the account until seven hours ago, and all the tweets after that point are spam.
Several options:
* Spammers creating accounts and posting human posts in advance knowing they'll be used for spam in the future
* Account sold by the actual owner
* Account compromised and repurposed for spamming
So within the next 10 minutes, teammates who will be leaving the company will receive a calendar invitation to a video call with a leader in your function and a teammate from the People team. In that meeting, you’ll review the details of your separation and have an opportunity to ask any questions you may have.
I would have sent the invite out beforehand to avoid the entire companying spending the next ten minutes in complete uncertainty about their futures.
Wow. AI is now good enough that writers are considered robot until proven human.
I don't think this will make much of a difference aside from the branding advantage. The quality of a mass-produced product like a consumer hand tool isn't a function of the country in which it is made but rather of its design. All the decisions made during that process and all the cost / performance considerations made are what determine the ultimate outcome.
It's common to blame Chinese manufacturing for cheap tools, but this gets the causation backwards. We wanted cheap tools and sending production overseas was the only way to get them. In fact because less is spent on labor, a better quality product can be had for the same price.
So bringing it back here without increasing the cost to the consumer means altering the design to make it cheaper to build, or banking on supply chain efficiencies to make it worthwhile.
Fortunately this is opt-in. Imagine being that guy managing a fleet of these things mowing a giant golf course and three times a day just stop for no reason and start singing.