My guess is that they can't object to foreign intelligence, and would lose negotiating ground if they even tried.
Optimistically, they can still refuse to do work that would aid in foreign intelligence gathering, by arguing that it would also be beneficial for domestic mass surveillance.
I'll admit that the phrase "We support...foreign intelligence and counterintelligence" is awful as hell, and it's possible that my apologist claims are BS. But Anthropic has very little leverage here (despite having a signed contract and so legally fully in the right), so I could see why they're desperate to stick to only the most solid objections available.
It's the addition of the we support phrase in particular, and the attempt to tie that in a "democratic values" clause that is objectionable.
Not to most US citizens, I'm sure. But there's millions of non-Americans who have given them their hard earned cash. It's not a good look, and it did not need to be phrased that way as it substantially undermines the impact of their point.
This is a public note, but directed at the current administration, so reading it as a description of what is or is not moral is completely missing the point. This note is saying (1) we refuse to be used in this way, and (2) we are going to use "mass surveillance of US citizens" as our defensive line because it is at least backed by Constitutional arguments. Those same arguments ought to apply more broadly, but attempts to use them that way have already been trampled on and so would only weaken the arguments as a defense.
If it helps: refusing to tune Claude for domestic surveillance will also enable refusing to do the same for other surveillance, because they can make the honest argument that most things you'd do to improve Claude for any mass surveillance will also assist in domestic mass surveillance.
Yes. As long as we're looking for relatively easy or cheap improvements, I believe that UX is a huge one. Buses have a long tradition of user-hostile design. "Exact change only", unhelpful and condescending and impatient drivers, unwritten etiquette rules, and everything else you listed.
It has always baffled me why they make it so hard for first-time users in particular. Sure, they mostly care about the regular customers who make up 99% of their passengers, but everyone has to be a first-timer before they can be a long-timer. It's not just UX papercuts, the experience seems designed to be maximally hostile. Is it because one more marginal person is a little more delay, a little more crowding, etc? It feels like there are perverse incentives at work.
> Buses have a long tradition of user-hostile design. "Exact change only"...
On every pay-in-cabin bus I've ever ridden, this is synonymous with "No change given". The machines are quite happy to accept more money than is needed for a single ticket, and the reason for that is pretty obvious.
> It has always baffled me why they make it so hard for first-time users in particular.
The SFMTA (the San Francisco bus/train operator) provides a document that addresses almost everything you brought up. [0] The "unhelpful and condescending and impatient drivers" thing isn't addressed, but I've never run into a Muni driver that was anything but helpful. [3] As an added bonus, the most useful information about fares is posted on the paybox inside the bus.
>> Buses have a long tradition of user-hostile design. "Exact change only"...
> On every pay-in-cabin bus I've ever ridden, this is synonymous with "No change given". The machines are quite happy to accept more money than is needed for a single ticket, and the reason for that is pretty obvious....the most useful information about fares is posted on the paybox inside the bus
That's fair, but (1) when I was a kid and starting out riding a bus, I didn't know that; and (2) as that same kid, neither my family nor I had very much money at all and paying "extra" for something is just not something you do. Consider it a cultural thing. "inside the bus" is good but insufficient when I'm deciding between walking a mile or chancing the bus that I don't understand. (I almost always walked the mile. I was cheap, and I hated looking stupid in front of unsympathetic people.)
As for Muni, I didn't live where I could use it until I was no longer that kid. But adult me fully agrees with you. My experience with Muni has been much better than with most other busses I've used.
I see. Your complaint is that in vehicles that are staffed only with a driver, the driver refuses to handle change, and that -in your youth- your parents didn't provide you with any information (whether directly from them, or published by your local transit authority) about how mass transit worked in your area.
There's not much the transit authority can do about your parents' decision to leave you ill-informed. I can tell you that obligating the solo driver to handle change would be significantly user-hostile for the passengers currently on the vehicle. The tradeoff made is the correct one.
As you're probably aware, there's also good news: for a while now, many (most?) transit systems permit payment with radio cards that are linked to a preexisting pool of money, rather than having to handle cash inside the vehicle.
> I believe the central thesis of this article is unsupported, and other assertions are false.
...but then your "In my experience..." section repeats the article's assertions? As in, everything you list as a drawback of riding the bus is exactly what the article claims can be improved by intelligently cutting out some percentage of stops.
Also, I didn't see the claim that "too many stops is the main cause of low ridership." That would be an overreach. The central claim that I see is that optimizing the number of stops, which turns out to result in a net reduction in pretty much all major American cities, is a relatively easy way to marginally improve many aspects of bus systems.
I think your counterarguments are valid, but they're just fleshing out the article's thesis. Simply reducing bus stops and holding everything else constant would not magically improve ridership and the overall experience. And as you say, reducing bus stops and removing money supporting the system will definitely not result in improvements. (And I agree that it is the likely way it would transpire politically.)
You would need to reduce stops and direct the savings into improving the remaining stops. You would need to convert the change into more reliable schedules. To make sense, that would need to increase ridership, and adjust the demographics of riders to include people who don't have to accept "dirty, unsafe and hostile" because they have no other choice. There's little incentive to improve things when the audience is captive and powerless. Also, increased ridership leads to more resources to accomplish the rest.
Of course, the dependencies between all these changes make the improvements more speculative and harder to achieve politically, so I do agree that you can't "just" reduce the number of stops and improve everything. As you said, that would more likely just drain more blood from an already anemic system. But the article is talking about a relatively cheap and easy way to improve things; everything else transit agencies can do is harder and/or more expensive.
> So while the number of bus stops might matter at the margins, we’re not talking about a system where marginal improvements will matter. If you want to improve ridership, you need to make the bus an attractive option for more people.
That first sentence says marginal improvements won't matter. The second sentence says that marginal improvements ("an attractive option for more people") are what are needed. Maybe you're saying that marginal improvements have to reach a threshold in order to be worth doing or achieve any noticeable gains?
I guess this is Anthropic's DRM moment. (Mozilla resisted allowing Firefox to play DRM- limited media for a long time, until it finally had to give in to stay relevant.)
I don't know enough to evaluate this or other decisions. I'm just glad someone is trying to care, because the default in today's world is to aggressively reject the larger picture in favor of more more more. I don't know how effective Anthropic's attempts to maintain some level of responsibility can be, but they've at least convinced me that they're trying. In the same way that OpenAI, for example, have largely convinced me that they're not. (Neither of those evaluations is absolute; OpenAI could be much worse than it is.)
My guess? Require them to not do the reinforcement learning on a custom model that implements guardrails. I think Anthropic has some of this built in already and couldn't alter it without retraining, but there's tons more layered on top.
> Also, the Government money would be a nice bonus, of course, but basically this is an existential threat for Anthropic.
It's also an existential risk to them if they cave in. What is the point of the company's existence if it's just another immoral OpenAI clone? May as well merge the companies for efficiency.
It's outrageous that the government is using the "supply chain risk" threat as a negotiating tactic. I know, I know, for the current administration it's unsurprising, but this is straightforward abuse of authority. There is no defensible claim that using Anthropic is a risk to anyone not trying to use it for murder or surveillance. At worst, it could be seen as less effective for some purpose, but that is not what "supply chain risk" means.
Could be challenged in court? As in, could a challenge win?
Horrible stuff is happening every day, so outrage fatigue is real. Still, try not to normalize it. Explain to yourself exactly why something is or is not a problem, before moving on to attempt to live your life.
> During the conversation, Dario expressed appreciation for the Department’s work and thanked the Secretary for his service
Ouch, I wonder how he rationalized that "service" part. Maybe by internally rewriting it to "thank you for all the positive things you have done in your position so far"? The empty set is rhetorically convenient.
reply