wood_spirit

Just curious, did the move on Greenland and Iran have national security interest? And is it economical or life threatening?

The general problem the average user has with a metered instead of provisioned billing model for computer services is you can’t easily control for cost overruns. From the old days customers getting stung for hosting costs when slashdotted or DOSed, to last decades microservice shock horror of the CI retry loop that burns money overnight to today’s AI that you basically have no idea how efficient the AI will be while it ponders your question, you are just setting yourself up for disappointment and cost overruns and a feeling that you’re not getting the value for money you got last week etc.

Yes air source are really good value in cost effectiveness terms, especially when a house has an existing central heating system they can connect to. But their COP - whilst dramatically improving in the last 10-20 years - is still behind ground source, particularly in the north during winter

The downvotes are unfair.

The price of things - heat pumps and alternatives - in different regions - even different regions within the US - varies by what people are prepared to pay not what they cost to produce.

The nordics have traditionally had cheap heat pumps whereas piped gas is only in the biggest cities and I’ve never seen bottled gas in the countryside. The competitor used to be cheap electricity and wood. Ground source heat pumps for rural install have been priced to compete with wood.

In the US the market could be shaped by regulation and taxation etc. It’s the choice of the US to have cheap fossil fuels and not embrace tech instead.

In the nordics it is common to have ground source heat pumps (brine in closed circuit pipe or bore hole) that are run backwards in summer to cool the house while actually assisting in storing heat back in the ground to extract in the winter. It’s a bit like regenerative breaking on electric cars.

The examiners would delight in you entertaining them. It’s fine to argue anything, wilfully misinterpret the question or say outlandish things. All that matters is you give the examiner a window into your mind showing it to be clever and articulate and rounded?

Yes that was one issue. It’s not the general degradation I have been talking about though, which is ongoing.

I recall reading similar tales of woe with other providers here on HN. I think the gradual dialling back of capability as capacity becomes strained as users pile on is part of the MO of all the big AI companies.

Me and so many coworkers have been struggling with a big cognitive decline in Claude over the last two months. 4.5 was useful and 4.6 was great. I had my own little benchmark and 4.5 could just about keep track of a two way pointer merge loop whereas 4.6 managed a 3 way and the 1M context managed k-way. And this ability to track braids directly helped it understand real production code and make changes and be useful etc.

but then two months ago 4.6 started getting forgetful and making very dumb decisions and so on. Everyone started comparing notes and realising it wasn’t “just them”. And 4.7 isn’t much better and the last few weeks we keep having to battle the auto level of effort downgrade and so on. So much friction as you think “that was dumb” and have to go check the settings again and see there has been some silent downgrade.

We all miss the early days of 4.6, which just show you can have a good useful model. LLMs can be really powerful but in delivering it to the mass market Anthropic throttle and downgrade it to not useful.

My thinking is that soon deepseek reaches the more-than-good-enough 4.6+ level and everyone can get off the Claude pay-more-for-less trajectory. We don’t need much more than we’ve already had a glimpse of and now know is possible. We just need it in our control and provisioned not metered so we can depend upon it.

Reminds me of Roy Van Rijn’s prototype that got a cease and desist letter! Lots of community disappointment at the time!

https://hn.algolia.com/?q=royvanrijn

I write bad, but my text editor is putting little grammar and spelling squiggly lines under everything and I click through them and end up with very AI-like text. My emails even end up with emdash in them. It’s to shrug. You don’t know if text today is completely prompted or is just cleaned up by modern grammar and spell checkers?

How does a highly automated factory zone in the Philippines gel with the administrations promise of bringing manufacturing jobs back to the US?

Such a disconnect from the minutes I’ve lost and given up on Gemini trying to get it to update a diagram in a slide today. The one shot joke stuff is great but trying to say “that is close but just make this small change” seems impossible. It’s the gap between toy and tool.

Most uranium mining is from Russia/CIS and those African counties that have experienced the recent wave of Wagner-assisted coups. The West needs to be energy independent, not just swap who it is dependent upon?

Through a modern lenses and with hindsight and memes that follow from napoleons failed attempt we can say flippant things like this and say napoleon wasn’t a genius. But students of history and future generals will obviously investigate the facts and see things through the lenses and circumstances of the time rather than from the seat or the internet commentator.

And Swedish houses, particularly detached houses built or renovated the 70s. Typically used for storing boxes.

Or the company data has been compromised. That’s a really common way for emails to ‘leak’.

I remember being fascinated by animals as a kid and the mystery of where eels went was one of the big unsolved puzzles I remember hearing about.

Much more recently I heard on QI about how medieval people, not knowing about migration, believed, through a lot of leaps, that it was ok to eat barnacle geese at lent. Worth investigating if you are curious :)

Great read, thanks :)

This is the kind of buzz I search out in my own programming :)

Have fun and keep challenged :)

Thanks for making and thanks for sharing :)

I’m not a parallels kind of user but I can appreciate your craft and know how rewarding these odysseys can be :)

What was the biggest “aha” moment when you worked how things interlock or you needed to make both change an and b at the same time, as either on their own slowed it down? Etc. And what is the single biggest impacting design choice?

And if you’re objective, what could be done to other tools to make them competitive?

I found sonnet 4.5 struggled with a two pointer interval merge (two sorted lists of things with start stop timestamps), but opus 4.5 managed. Then it took opus 4.6 to make it a three way or k way merge. So it reminded me of the classic Simple made Easy talk by Rich Hickey where he talks about braids. Sonnet couldn’t track two twisted braids and opus 4.6 managed a weave.

But the last few weeks Opus 4.6 seems to have got dumb again. Now it is making way more mistakes and forgetting useful things and recent context it used to manage.

I am guessing this is just Anthropic quietly dialling down the real effort as they either downsize or free up compute for someone or something else.

There should be rooms assigned for these private conversations.

But I imagine even these rooms are cammed, and lip reading is a thing

Not a rubyist so just curious on the background and if this is the “good” or the “bad” side in the spat? What’s the other side and what has been the broader community impact?

The company has an internal policy of “open company, no bullshit” and an internal channel for venting called literally “outrage”. I don’t see an “official internal” and “unofficial internal” distinction here.

And report back please! :)

(Perhaps a good library for timestamp code in data pipelines https://github.com/williame/TimeMillis)

Also zap the timestamp instant objects if you really need speed; see https://github.com/williame/TimeMillis

Yes! Obligatory link to the seminal work on the subject:

https://gwern.net/doc/cs/2005-09-30-smith-whyihateframeworks...

A subject close to my heart, I write a lot of heavily optimised code including a lot of hot data pipelines in Java.

And aside from algorithms, it usually comes down to avoiding memory allocations.

I have my go-to zero-alloc grpc and parquet and json and time libs etc and they make everything fast.

It’s mostly how idiomatic Java uses objects for everything that makes it slow overall.

But eventually after making a JVM app that keeps data in something like data frames etc and feels a long way from J2EE beans you can finally bump up against the limits that only c/c++/rust/etc can get you past.

Does this have corresponding speed ups or memory gains for normal CPUs too? Just thinking about all the cups of coffee that have been made and drunk while scikit-learn kmeans chugs through a notebook :)