HN user
I don’t quite agree. Good human artist can visualize in their mind how to draw a picture, i think. Which i think is no different than LLM doing SVG drawing in their “head”. Anthropic’s recent post call this head-space “workspace”.
It just might feel foreign to human who does not have a SVG trained head-space.
I imagine all vision models have to do this, this being html rendering, to be able to do well in web design.
Why not? Look at the silicon valley AI researcher salary. There’s no way Chinese researchers make close to the same amount.
It’s CXMT (memory) not SMIC (logic)
Except that dynamo is still just glorified mysql? https://news.ycombinator.com/item?id=18871661
I don’t think the backend matters. It’s the frontend wrapper that makes or breaks HA.
It’s 2026, why can’t credit card and merchant figure out a way to transmit order summary URL as part of credit card transactions so I don’t need to match up transactions by amount??
Maybe author is writing this from an Intel Mac!
It’s also interesting that Cursor (that spacex is acquiring ) also just announced yesterday they are training in colossus.
So who’s using it? Is spacex just renting out parts of their data center? Or is cursor done done?
They better do. What is google doing?
The interesting thing I find is how Anthropic has been more consistently improving over time in the last few years, that allows it to catchup and surpass OpenAI and Google. The latter two have pretty much plateau over the last year or so. GPT 5.5 is somehow not moving the needle at all.
I hope to see the other labs can bring back competition soon!
Does it not depend on what the feature is?
Is it theft if a company stops supporting TLS 1.0 clients that were previously supported?
What about upgrading service to only support a new form of authentication, breaking some clients?
What if the new authentication can only be done by locked down official client?
What if the company never advertised support through third party clients?
Before: user can print via cloud or locally with custom slicer at the same time easier.
Some time in 2025: firmware updates make user choose between cloud XOR locally. Enabling local mode allows using custom slicer, but disables cloud printing or monitoring. Folks were up in arms because they wanted both, and openness.
Latest fork: a specific new custom slicer impersonates UA to submit print via bambu cloud, so it gives the pre-2025 experience.
Bambu sues this new fork. Actual OrcaSlicer working locally is fine.
Is it really failing more, or we just don’t hear about failure happening elsewhere?
Last i heard azure outage it wasn’t even on HN frontpage
That seems fine as long as they can show lower volatility than market while still being close in return?
Did they?
LLM are a text prediction engine. Starting the prompt with “you are a helpful assistant” help make subsequent text prediction more in line of that of a helpful assistant.
It’s like the law of big numbers. Once a project grows large enough, some entitled free-riders are bound to pop up.
What to do as maintainer? Can everyone of them find piece?
Having a SSD certainly is better than no SSD.
We don’t know if Palantir is using claude for those uses. Though anthropic would not know for sure either.
I do agree with your point that Amodei is playing a game though. Whether he’s winning the bigger picture or not it’s unclear. His red lines are already so watered out, like how domestic surveillance is not ok, but international? totally fine.
It’s like money laundering, but now responsibility laundering.
Anthropic released Claude saying “hey be careful. But now that enables the masses to build OpenClaw and go “hold my bear”. Now the masses people using OpenClaw had no idea what responsibility they should hold.
I think eventually we will have laws like “you are responsible for your AI’s work”. Much like how driver is (often) responsible for car crashes, not the car companies.
Yes larger format TV is getting more affordable, but I don’t think larger living room is catching up. Watching tv in general also feels like a dying trend. I would not be bullish on larger tv getting popular outside niche enthusiasts market (with money to buy a mansion)
What model did you use?
At work tasks that Sonnet 4 and 4.5 failed miserably, Opus 4.5 can basically one shot them. I imagine it will be the same here.
I agree. It is basically claude code running dangerously all the time. That is actually how I use CC most of the time, but I do trust Anthropic more than random github repo.
(I have the same sentiment about manifest v3 and adblocker, but somehow HN groupthink is very different there than here)
Edit: imagine cowork was released like this. HN would go NUTS.
The whole world is about bundling (and unbundling).
Not saying it really is useful, but there are values bundling an easier interface to CC with battery included.
AI PC has been in the buzz for more than 2 years now (despite itself being a near useless concept), and intel has like 75% marketshare for laptop. Both of those are well with in norm for an intel marketing piece.
It’s not really meant for consumer. Who would even visit newsroom.intel.com?
Because that's not how they perceive their works. Instead it is "advocating for one's own team and passion", "helping others advance their career", "networking and building long-term connections".
The west is already ahead on this. It is called AI safety and alignment.
Where’s the source for this?
It doesn’t look good when similar WAF issues caused their big outage a few years back.
I just bought the same and planning to break out just the touch id to put into a 3D printed enclosure. Expensive hobby…
GPT4.5 was allegedly such a pre-train. It just didn’t perform good enough to announce and product it as such.
Allegedly deepseek is doing this because they don’t have enough gpu to serve two models concurrently.