nl

I agree that OpenAI's motive is profit.

I stand by my moral position.

This is no longer the case.

Opus and Sol usage levels vs the API are currently roughly the same, but Opus 5.5 outperforms at low and medium effort levels.

This is completely the wrong read!

Sol 6.1 scores one point less than Gemini 4 on intelligence AND costs less than half ($0.72 vs $1.99) per task.

Additionally, if you are using OpenAI you have the option to pay a bit more and get Astra which - despite the benchmarks - does outperform Sol on some things.

Also, people are - rightly - very wary of Google's benchmaxxing tendencies. I think lots of people remember Gemini 3.0 (I think?) which benchmarked amazingly, but as soon as you used it would go off-track and needed constant babysitting if you wanted to use it for agentic work.

"Auto" mode in Claude works really well if you don't want to do dangerously skip permissions.

This was a problem maybe 6 months ago. With Fable/Astra orchestrating Opus/Sol subagents it now works really well.

Ask it to interview you before you start and it'll get your constraints pretty well.

Iran is mostly self-sufficient in food. I don't think there is social license in the West to starve a civilian population.

> So who is more on the clock? I guess time will tell.

Well we know the US's deadlines. The midterms are one, and in 2 years time US policy will change (although unclear how exactly).

Yes it's very painful for Iran, but when have economic sanctions ever created regime change (Or whatever the US's actual goal is)?

The Iranian expats I talk to (who are all very very opposed to the Iranian regime) are depressed because their read is that the regime has increased support within Iran because of their success standing up to the US.

The sanctions have made Iran much closer to Russia and China though.

> In the end Iran's goal seems more attainable: Make this so painful for the US that they do not dare try this again for a long time if ever.

> The US's goal was topple the regime, install their puppet and acquire the nuclear material. Now its just getting back to the JCPOA or something like it. Not good enough for Iran. They want to make sure this stops for good.

I agree with this analysis. If the US could walk away with JCPOA back they'd love it (of course they couldn't call it the "Obama Iran deal" anymore, but putting that aside).

Iran seems to think they can get a more favorable deal because the US has no appetite to make this a long war. I think there is evidence to support that position.

This is not true. Don't make things up.

In 2015 the SDR was authorized to sell $2 billion worth of oil during FY2017–2020 and use the proceeds for infrastructure upgrades and life-extension work at the reserve's aging storage facilities.

There were also sales between 5 and 10M barrels a year to fund the "21st Century Cures Act" (https://www.npr.org/sections/health-shots/2016/11/25/5031763...)

The background to this was the increased domestic oil production which made the SDA seem less necessary.

https://www.eia.gov/todayinenergy/detail.php?id=29692

The Biden administration (and other countries around the world in the International Energy Agency) released reserves to try to stabilize the spikes in prices due to the Russian invasion of Ukraine.

> On March 1, 2022, Secretary Granholm announced that the U.S. and its partners in the International Energy Agency (IEA) would release a total of 60 million barrels of oil onto the world market. [snip] In total, DOE released 180 million barrels of crude oil into the market.

https://www.energy.gov/hgeo/opr/history-spr-releases (look at 2022 IEA Coordinate Release)

I think the general interpretation that both sides think something will change after the midterms is reasonable.

But

> Iran et al are trying to hold out until the election

> [Iran will] throw in the towel as they lack the ability to hold out long term

is something that I don't think even the most thoughtless parts of the administration really think right?

It's obvious that Iran is winning, not close to throwing in the towel. The US is reduced to trying to get the Straights of Hormuz open without paying a toll to Iran, which is a worse position than the US was in at the start of the war.

I think you need to do some reading on externalities and how they lead to misallocation of resources in a free market.

I think this is a fair point, except that in this case the same laws[1] are being applied to both, and the person I was replying to was literally arguing both ways!

[1] broadly the same laws: I do understand one was criminal and one is civil and yes I agree that injustice. To me neither should be criminal.

I think it's good when eight-hundred and fifty-two billion dollar, for-profit corporation establish legal precedent than can then be used to protect activists.

I agree entirely with this position!

But it is actually possible to agree with one thing a company does and disagree with others.

> Aaron Swartz and the fate he suffered?

I think what Swartz did was moral, and his prosecution was unjust.

I think what OpenAI did was moral, and them getting sued for it is unjust.

Why do you have one position for Swartz and a different one for OpenAI?

(Aaron Swartz was a mailing-list friend of mine, so I do have some bias here. But in part we knew each other because our moral position on this was similar)

I've always wanted to play Diplomacy but never have. Is there a matching service or something?

Oh I agree that "know your enemy" is good advice!

But an alternative commentary with the same facts would have been something like "Russia knew there was wide Western support for Ukraine but thought they would overrun the country before Western nations would be able to provide practical support. They dramatically miscalculated and now are trying to deter Western countries by slowly increasing actions."

This was basically what @jwr said that kicked off this subthread:

> It's a typical Russian strategy: gradually increase the intensity of operations, while pretending you have nothing to do with it.

Read the context of their comment:

>> It's a typical Russian strategy: gradually increase the intensity of operations, while pretending you have nothing to do with it. That is exactly how war in Ukraine began (way back with the first incursions of "little green men").

> But their [Russia's] problem is that the equivalent weapons factories, refineries etc. often aren't in Ukraine, but in NATO countries supplying Ukraine and giving Ukraine explicit permission to use them against targets inside Russia

That reads exactly like an argument justifying Russia's actions.

Yes, there is no doubt that NATO countries (except maybe the US/Trump some days) wants Ukraine to win. They train troops, help with intelligence etc.

There is no argument that they are neutral.

But there's a big distance between being an adviser and being at war.

North Korea is supplying actual troops to Russia, and that doesn't mean Ukraine is at war with North Korea. Belarus and Iran have both supplied arms to Russia and Ukraine isn't at war with either.

Palintir and Starlink are both private companies, not NATO.

Have you tried it? It's pretty good!

International reporting was almost all that it was a Ukrainian operation, eg:

> On 6 June 2023, The Washington Post reported that an intelligence agency of an unnamed European "close ally" had notified the CIA of a Ukrainian military plan for a covert attack on the pipeline. The plan was to be carried out by a team of divers reporting to Gen. Valery Zaluzhny, commander-in-chief of the Ukrainian armed forces, but the Post did not report any evidence that it had been put into action. The US reportedly learned of the plan three months before the destruction of the pipeline, and shared the intelligence with Germany and other allies.

https://en.wikipedia.org/wiki/Nord_Stream_pipelines_sabotage...

No, they are trying to justify actions against countries that supply arms to Ukraine.

As I point out in a other thread, supplying weapons is pretty normal and attacking random infrastructure in a country that happens to be allied to Ukraine is a significant escalation.

Countries have always had arms supplied from countries not involved in the conflict.

It's a significant step to escalate to attacking those countries.

> Russia views it as NATOs proxy war against them now

It is an interesting argument to choose to invade a country, act with faux outrage when they find allies willing to supply them with arms and then call it a proxy war.

Russia got themselves into this mess. They have an easy way to get out of it.

Extremely. It's been top of the App store charts since it came out in the US.

5.5 is higher for max effort, slightly higher for xhigh and lower for high, medium and low effort.

The biggest proportional difference seems to be at max (5.5 is 38% more) and at high (5.5 is 21% less).

I think most people run at high and xhigh. At xhigh it is close enough to be task dependent and I don't think most people will notice. At high effort I think it looks like it will be an improvement for most people.

5.5 Max should probably be compared to Fable - it performs a lot better than 5 Max.

https://artificialanalysis.ai/models/claude-opus-5-5?models=...

Originally it was because Anthropic was so compute constrained they relied on the extra care the Claude harness took with caching (heavy use of cache breakpoints etc) that other harnesses didn't.

I think that is less of a factor now, and I think Anthropic have backed off some on being as strict (eg, AFAIK they never implemented the two-tier "claude -p" pricing model they were planning)

Oracle sued Google over Java - it wasn't some previously existing case.

Yes, it is Oracle so there's a great chance they were considering this when they bought it.

I believe most speculation is that Opus is 5T model, Fable a 10T

Elon claimed Opus was 5T in April, and I think it's fairly likely this is accurate: https://x.com/elonmusk/status/2042123561666855235

This isn't necessarily true.

While every formal proof can be encoded mathematically that doesn't imply that the grammar of lean or rocq is sufficient to encode every potential proof.

However the core of what you are saying: that every possible proof exists in the space of all mathematical statements is correct.

> The MySQL purchase. WTF, this was horrible and stupid. The only interesting effect of this was to make Sun a target of acquisition for Oracle.

Strong disagree with this!

The MySQL purchase wasn't a lot of money and was a move towards becoming a systems company (like you suggested they need to be)

Oracle bought Sun for Java, not for MySQL.