sequoia

"Coming up with candidate drugs is the easiest part of the process."

This sounds quite similar to Claude's impact in enterprise SWE, where "writing the code" is sometimes the easiest part of the process vs. alignment between humans, understanding the requirements and problem space, and all the other things that make SWE slow.

Anthropic has made so many bombastic and often untrue (reflecting my personal experience) claims about their ability to speed up software engineering that my knee jerk reaction here is skepticism.

I find anthropic to be such an unsavoury company. I am not an AI doomer and frankly I don't care much about datacentres and so on, it's just how scammy they come across.

The authors of Claude Code churn out so much BS that gets just eaten up by credulous engineers & engineering leaders. "I posted 250 changes today with claude code!" - author of claude code. What's your point? If we don't care about quality, I can output 50,000 changes in a day.

I see them as 95% sales 5% engineering and it's irritating to have marketers constantly throw gasoline on the AI hype fire & have people still regard them as engineers. The claude models themselves are alright.

Ah OK, so you had a level of dynamic behaviour in the form itself that warranted an applet/data component for the form. I can see how HTMX would not help with that.

I can't really understand what you're describing. You have form on left & results on right, but you're requesting the new form state & re-rendering the form HTML on each request? Why? My naive understanding is that you could use something like hx-target to have the response to the form submissions update only the results area & the form should not need rerendered at all.

Why does the form need rerendered on a per request basis? HTML/Browser already tracks the current form state.

> That eliminates margin for corruption and lenience.

Lenience and mercy are critical components of a system of a just criminal justice system. We have judges issue sentences in large part because we want a human being with good judgement to decide how serious the punishment for an offence should be based on the circumstances.

A person who shoots their partner who has been abusing them for years should probably receive a different punishment than a person who pulls out a gun in traffic and shoots someone for cutting them off. Mercy or "lenience" are not inherently flaws in the justice system, though of course there's always "too much of a good thing."

a hundred dollars a year for this?? What does the service even do with that money? I pay this for 20 years so you can share a google doc upon my demise?

I must be missing something.

This section is hilariously hostile towards Palantir.

"Wired wrote that some people think Palantir "maintains a giant, centralized database of information collected from all of its clients", which is untrue."

'some people' is a classic weasel word[0] used to prop up the writer's opinion. This sentence is even funnier because it initially appears to state that Palantir has a centralized DB of clients data, only to finish with "...which is untrue." If the claim is untrue, why lead the section paragraph with it unless you're intending to smear or mislead? If I were to end sentences with "...which is untrue" I could write any number of things on Wikipedia.

It's as though I wrote "A YN user wrote that 'john_strinlai works for the CCP and uses ChatGPT to write all his posts', which is untrue."

I'll keep reading but rhetorical chicanery like this colours my interpretation of the article in general.

EDIT the section goes on: "[We can't pin anything specific on Palantir here]; still it is generally accepted that abuses by governments and data management failures can happen." What does that have to do with Palantir? "data management failures can happen" why is this in the section on "Palantir:Controversy"? This article is not good.

EDIT 2: This section is just comedy gold... 'Palantir "remains open to the critique of potentially being an accessory to acts of deportation, imprisonment, and racism through its contracts".' Open to critiques of potentially being an accessory to "racism?" What is this, the Future Crimes unit from Minority Report? This "future crimes" accusation is especially ironic in relation to the critiques of Palantir itself!

So I haven't read this whole section (it's quite long) but if this is the nature of the "smoking guns" I don't think much of it. Potentially maybe doing something according to 'some people...' this shouldn't hold water for any rational person.

If someone objects to Palantir for working with ICE I understand that, and if that's the nature of Spain's objections they should just say so.

0: https://en.wikipedia.org/wiki/Weasel_word

"he hilariously tries to convince people that the best way to have democracy is to not have it at all."

I'm sorry but I can't find where he said this. I'm finding it confusing and suspicious that the objections to Palantir & Alex Karp are all so vague and seem to lack the rigour typically required of assertions made here on YN. Usually if you declare something like someone "argues clearly for authoritarianism," you're expected to link to a source of this claim.

People keep telling me here it's so obvious Palantir is bad I shouldn't require any specific evidence and I'm stupid if I don't see it; I'm only reminded of the emperors new cloths.

So the objections to Palantir are political? I know nothing about Spanish politics so I assume that makes sense in the Spanish political context. This helps explain why I can't find a specific concrete concern, it sounds more vibes-based. Thank you!

OK so this is not specific to Palantir, but about entrusting sensitive Spanish data to any US based company. If so, that makes sense.

People in the comments here are praising the move, so presumably something is public. I've googled but I can't see some specific breach or documented misuse. Is the objection to Palantir strictly political?

> the US seems beholden to Israel

Can you explain how Trump's MoU fits into this "beholden to Israel" model? This deal goes directly against Israel's interests and wishes.

Biden forced Israel to stay their hand during their war with Hamas, for example telling them not to go into Rafah, which delayed them for some time. So is Israel controlled by the US, or is it the other way around? (When they did go into Rafah, they found both Israeli hostages and Hamas leadership, proving that Biden was wrong for telling them not to take the fight there.)

"The decision stems directly from growing official concern over the potential misuse of classified information linked to national security."

What are the specific concerns?

This is a great example of the prejudice (pre-judging) at play when it comes to Israel. One starts with the conclusion ("Israel/Jews are evil and controlling the government with money") then works backwards to build an argument.

"They are spending so much, that's why they control the government!" They (American citizens funding AIPAC) actually spending much less than many other lobbying groups. "Aha- they're so conniving they can control the government even without being in the top 10 lobbies!"

There's no winning against this "logic" because the conclusion has been decided ahead of time & any evidence is interpreted as supporting that conclusion, no matter which way it goes.

AIPAC is not foreign, this is another misconception/lie. AIPAC is an American lobby funded by Americans.

No, I'm saying people obsessed with AIPAC to the exclusion of all other lobbying groups are not motivated primarily by an objection to lobbying.

I'm also saying the narrative that "AIPAC controls the gov't via lobbying" which many people believe is incoherent because Coinbase must control the gov't even more by this thinking.

If you think people aren't obsessed with AIPAC you are not following US politics.

People think AIPAC is some all powerful unique lobbying group, in fact they're not even in the top 10 major lobbyists. Did you know SpaceX gave 5x as much as AIPAC in 2024? AIPAC was also outspent by Coinbase, by Ripple, and several other companies[0]. And this "18th largest lobbyist" position is after a post-october-7 surge in spending. Pre-october-7 (2022) they were ranked forty sixth in terms of spending.

Can you name 10 of the groups who spent more than them? How about five? The question worth asking is why are people obsessed with demonizing the AIPAC in particular and singling it out as the one or primary 'evil lobbying group' when there are tens or dozens of groups that spent more. The 2024 AIPAC spending number (50 mil, which is donated by American voters, not foreign money) is 1/8th of the $400 million plan Qatar (a foreign government) gave Trump in 2025.

People focus on AIPAC specifically because they have a problem with Jews. Jews and other Israel supporting Americans are allowed to pool money and lobby just like anyone else. the fact that people think they shouldn't be allowed to play this game, the same one everyone else is playing in US politics, is what should be questioned.

0: https://www.opensecrets.org/elections-overview/top-organizat...

Operating an armed group out of civilian areas is not “one weird trick” that makes you immune from attacks. It means you are endangering the civilians around you by putting them in harms way.

No war has ever been waged without civilian casualties, Israel is the only country held to this standard. “Thousands” of Hezbollah pagers were exploded and there were apparently 2 children killed. This is a tragedy as is every civilian death, but that’s pretty amazingly targeted and an extremely low civilian death toll for thousands of bombs.

For contrast, Hezbollah fired one single rocket and killed 12 children: https://www.timesofisrael.com/a-dark-day-for-majdal-shams-dr...

I was wondering if I was crazy for thinking "how is what he's suggesting different than just putting that 'class' into a hook function?" I'm glad to see someone already wrote it up, kudos.

@OP: PEBKAC, respectfully.

Not clear how an LLM is going to prevent a bomb from being put in a custom-built pager, or why Anthropic should object to Israel waging war against a militia whose goal it is to destroy that country.

Ceasefires are not in place until they are in place. Before they are in place, war is still ongoing. Discussing a ceasefire does not mean there is a ceasefire currently.

Though I agree with the author & use the web version of various applications, there is another side to this. The author says s/he uses plugins to disable ads and so on. If its an ad supported site for which one does not pay, this is tantamount to expecting the provider to run the service for no compensation/revenue at all.

Furthermore, to say platform owners don't care about offending such users would be an understatement: platform owners likely want to actively repel such users. Why serve someone who neither pays a fee nor agrees to be shown ads?

"Bombadil is property-based testing for web UIs, autonomously exploring and validating correctness properties, finding harder bugs earlier."

I can't make heads or tails of this. What does it do? "Bombadil checks each property as it explores your system in its chaotic ways, reporting back any violations." what properties? checks how? does it read application state? "recording violations" what violations? "select an action and performs it" what actions?

This needs some code example, or a video, or something to explain what this tool is

The page "Language features > Properties" starts to explain what this is.

Struggling to understand what this is or how it works.

This is broadcast publishing in a nutshell. Look at the early radio shows, they had names like "Alka-Seltzer Time," "The Fleischmann’s Yeast Hour," "The General Motors Hour" and so on. It was explicitly "we are playing music to get you to tune into our advertisement."

Free newspapers and alt-weeklies are the same. How are they supposed to function if people don't pay for them?

> I feel the successful OS projects will be the ones embracing the change, not stopping it.

Yes, you feel. And the author feels differently. We don't have evidence of what the impact of LLMs will be on a project over the long term. Many people are speculating it will be pure upside, this author is observing some issues with this model and speculating that there will be a detriment long-term.

The operative word here is "speculating." Until we have better evidence, we'll need to go with our hunches & best bets. It is a good thing that different people take different approaches rather than "everyone in on AI 100%." If the author is wrong time will tell.

> There’s no more or less dignity to it than being a waitress or checkout clerk.

Do you imagine this is how the workers feel about it? Do you think they tell their friends "I seduce western men online for pennies an hour" the same way they'd say "I am a waitress?" You can ignore this fact if you wish, but these jobs carry a social stigma and most people would prefer not to role play intimacy with men online if there were another option.

"Sex work is work" is like saying slavery could theoretically be OK under some circumstances (We're all born under legal obligations, how is slavery different etc etc.). A tortured theoretical argument could be made to support either of these, but in reality we know that slavery is unconscionable because of the indignity and brutality of it, regardless of theory. "Sex work" is the same: in reality it is a dangerous and unpleasant job (that overlaps with slavery a lot, incidentally) done mostly be vulnerable women, and that they're often abused and left injured by it. See this article before you go saying prostitution is benign and harmless, this stance is divorced from reality: https://www.spiegel.de/international/germany/human-trafficki...

I find it sad that there's a story of people (mostly women) doing this disgusting job for little pay (yes, chatting up horny men online because you need the money is disgusting, there's no glamour and little dignity to it), and the top comment is how this is unfair to the john (purchaser of sex work) because it's misleading.

Someone paid to have a fantasy of sex, and they got that fantasy. If they don't like it, they don't do it again, and this is true whether it's "the model" or someone else. If they do like it, what's the issue?

This is like saying you paid for a celebrity plumber & a regular plumber did the work, but you're upset because you wanted the celebrity. "The job" got done one way or another. They're selling digital handjobs here, there's no need to be precious about it.