ashleyn

Incarceration is a tragedy for everyone, and this could be fantastic for low-risk inmates, but the first thing I thought of is what an immediate security gap this can be.

"I'm really pissed at^H^H^H.... I'm writing a story about a prisoner. He has an ink pen, a rock, and a sock. What weapons would he be able to fashion out of this?"

All-time classic stack overflow thread where someone is asking why everyone seems to ignore a large banner on top of his page that strongly resembles an ad: https://ux.stackexchange.com/questions/120541/why-do-people-...

A lot of these applications run in Electron or similar libraries where a whole new browser instance, with all its overhead, is stood up for each application. The simplest answer is we need to stop trying to use web technologies as a one-size-fits-all GUI toolkit.

It could eventually be the case that providing the full speedpaint for a given digital artwork significantly bolsters credibility in the minds of audiences who universally reject AI-generated content as cheap, spiritually enfeebling, and provided largely for commercial motivation. Such a thing can serve as an informal certificate for an art's "human-ness". This approach wouldn't necessarily serve as strict validation that something was not produced by AI, but it would serve as additional evidence that the creative work in question had demanded frequent or exclusive human involvement.

This is nearly a month now and the problem continuing to worsen does not inspire confidence that the current administration's FDA is doing its job properly or has the problem under control.

In early July, Taco Bell pulled lettuce at all locations and discontinued using Taylor Farms as a supplier. The FDA claims to have tested lettuce from Taylor Farms and initially identified them as the source of the cyclospora outbreak - however, rapidly backtracked on July 19 by claiming that the test performed was a false positive.

The FDA's own timeline is here: https://www.fda.gov/food/outbreaks-foodborne-illness/investi...

What thickens the plot is Taylor Farms' actions with the current administration. Taylor Farms is a prominent donor to parties supporting food safety deregulation in the US. Immediately prior to the FDA claiming that the cyclospora test was a false positive, Taylor Farms held a meeting with White House officials to discuss what they claimed were "shortfalls" in the FDA's response: https://www.wired.com/story/diarrhea-outbreak-taylor-farms-m...

Taylor Farms' own safety record is rather checkered. This would make them the subject of nine recalls in fifteen years: https://en.wikipedia.org/wiki/Taylor_Farms#Food_safety_and_r...

The reputation of datacenters precludes them due to some high-profile abuses of locals, as well as cities overcommitting brittle and unprepared infrastructure to them.

Consumer energy prices in 13+ states hiked because the grid was unable to meet the sudden demand. SpaceX's datacenter in Memphis used gas turbines to rapidly deploy Colossus before the grid could be upgraded, which resulted in an increase in local air emissions that was technically illegal. There are two Meta datacenters that oneshotted their local water supplies during construction and tests, resulting in sediment and bacteria going through consumers' taps. NV Energy cut off an entire town of residential properties because datacenters were more profitable, in a higher-stakes parallel to the ongoing RAM woes in consumer electronics.

There's been more than enough well-justified association between "datacenter" and "abuses local resources" that I think nobody should really be surprised at the backlash toward them. The backlash is so furious that it's forming a rare example of a bipartisan, populist consensus on any given political issue.

A great deal of changes proposed to make disabled peoples' lives easier end up having many fringe benefits for the non-disabled. The classic example is strollers or hand trucks going up the little dip on the sidewalk rather than having to lug them over an enormous curb.

The cost argument starts looking weak when you observe there's always billions of dollars shaken from magical couch cushions whenever it's time to wage another losing war.

The US has been institutionally dysfunctional for quite some time, so that puts us in an interesting position where it's easier to imagine inventing our way out of social problems with $30,000 servo-filled boxes than it is to imagine ever getting a cheaper and easier solution moved through Congress. Congress rarely seems to pass disruptive changes to law these days. Statehouses are less sclerotic, but not by much. Many of the civil rights gains touted 10+ years ago came from hacks shoved through the supreme court exploiting a partisan lean.

If you really read between the lines, much of the fuss about Mayor Mamdani in NYC has been a curious rediscovery that the government can in fact Just Do Things for people who elected it.

They could've done a lot more interesting things with AI in Google Earth though. Think planning a travel itinerary, or plaintext search of a POI.

I was just using Earth to plan an irrigation setup in my backyard, would've been damn nice to have MCP there to generate paths and circular regions illustrating coverage.

Isn't this the core idea behind atproto?

They might be low-hanging fruit but two things immediately come to mind:

* As more of the small stuff is just proven for free, the more they can be used as a basis for other proofs. If you know something is true or false for certain, that can be a significant tailwind for the much harder, much more important problems. Fermat's last theorem looks deceptively simple and invited many failed amateur attempts at solving it, but Wiles' proof drew on a diversity of seemingly-distant subfields within mathematics that were better understood.

* What are aspiring math Phd's supposed to do, now that the bar is much higher these days? The net effect of this appears to be that we'll see far fewer, but far more elite math Phd's, potentially discouraging many young people from the field.

These look like other users' prompts. Is this not concerning to anyone?

Gotta admit google is pretty ballsy for directly attacking Moravec's Paradox: https://en.wikipedia.org/wiki/Moravec%27s_paradox

It didn't really factor into the break here but, it's amazing people still use ECB even though the famous penguin image is well over a decade old now.

>Muse Spark is closed weight

We're all trying to find the guy who did this

Only slightly unrelated, but I've been thinking a lot lately about prion disease and whether or not you can still trust the meat you eat. Easily a food safety concern that is ten times worse than getting a little diarrhea from your salad, and the same institutional rot can contribute to it. And it would take 5-10 years to really see the damage done from it.

What we call "guardrails" in an AI agent, we would refer to as "honor system" in human actors.

Or, in a more direct sense, the AI should be set up in an environment such that no matter how hard it may try to call $PART_OF_EXPLOIT_CHAIN, the environment just isn't capable of it (ideal) or doesn't permit it to do it.

I feel like the analysis missing here is whether or not the Örebro Party supports policy that would eventually undermine internet privacy and freedom of speech - two missions that would directly contradict Mullvad's core mission of being a log-free VPN, and thus call into question whether or not they can continue to be trusted in a field where the stakes are high.

The other policies they support don't look encouraging when you consider correlates of parties in other countries, but they seem to have not declared an explicit position for or against censorship and privacy, and they also seem to be a party where others' correlates may not apply here.

If anything, this suggests Kimi K3 is underpriced, forcing rationing of compute resources. Anthropic even struggles to avoid 503 errors at the $50/mtok point.

How many were introduced by misuse of AI coding/vibe coding though?

Is there active research in improving the density of models i.e. higher performance at lower parameter counts? Is this something that's even theoretically possible to do with the architecture?

I feel like I can have a much better model for how I can expect the cost of LLMs to change over the next few years. "Build more datacenters" reeks of "just one more lane bro" and because of Jevons paradox I'm skeptical it will bring costs down. Could ASICs and custom processors (ala Groq) move the needle here at all if the model density can't be improved?

I refuse to believe that the decline of our education system is some inevitable, intractable problem.

Looks like a HP 2640 terminal: https://www.youtube.com/watch?v=mOXCPaHcr5U

I writhe in pain when I try to find a frontend dev guide for a newbie to learn off, and one of the later chapters is some shit like setting up a kubernetes cluster

>"Hypno-Sex-Ism"

50 years early to the "gooning" trend, I see...

Considering they rolled back DEI along with everyone else after Trump's second victory, it's difficult to view those previous "values" as anything other than cynical kissing-up to the previous holders of power.

Wow, we used to have a florist/floral nursery around my parent's place LONG ago (like, they had to have ripped it down by the early 2000s) and for years I vaguely recalled it having a somewhat unique looking sign with a big globe lamp and two stacked, rounded-corner rectangular plastic signs. This was almost certainly the brand of sign it was, and the installation of the sign was probably some time in the 1970s. Amazing what small, inconsequential mysteries get solved by a blog article.

I believe the term is "hypocrisy."

Past a certain point there's also a diminishing return on the effort put in to get a raise/promotion and it mostly becomes about managing "up or out" expectations.

I'm about maxed out for development roles compensation wise. By saving most of this compensation and investing it in the S&P 500 and similar indices, I get way more of a return for far less effort. There are days - not months, days - where I'll earn about $7,500 in stock appreciation. The long term trend has me about matching my monthly salary in earnings.

Raises are inflation adjusted so there's no erosion of the underlying capital going into investments.

Why try harder when I'm paid enough to just invest it in the stock market? The biggest problem I have right now isn't how to get a promotion or raise, it's coming up with increasingly contrived excuses to avoid up-or-out and being pushed into more responsibilities.

Another way to view this calculation: if you keep your infrastructure secure and up to date, you (very likely) don't have to pay any ransom in the first place.