Which modern TV has good software?
Seems like pretty much all TVs today come with horrible software.
I'd very much like the option to only have a basic OS on there that's actually snappy like my old TVs used to be and not need to load its menu from the internet every time I start it.
They still need to choose when to do that though. When I prompt the program to e.g. alter a bash script in a specific way or to recite a longer known text it can't go round and randomly exchange tokens. It has to somehow define what is a simple repeated text from a different origin and what is a novel generation.
Personal observation: Opus 5, over the last week, has started outputting A LOT more comments. Despite my global instructions being full of variations on "don't use comments unless absolutely necessary".
I might be imagining things of course. But comments would be great fit for this use case.
Comments seem most plausible, especially since I absolutely expect it to match my code style, existing architecture, and have the code go through CSharpier and dotnet format after the fact.
edit: as an aside - I actually use extensions to collapse comments and change the color to be less intrusive.
Yes the length of comments Opus 5 leaves is exhausting. Not to mention it will insert info thats only relevant within the current session. I've just been deleting all of them lol
You are aware that Claude already doesn't choose the most probable token, right? That's literally what the temperature parameter means. It picks tokens at random (from the list of most probable next tokens), increasingly so as the temperature goes up. This has always been the case. And now, with the watermarking, it will simply go from random to pseudo-random, adding some patterning.
There isn't any effect on the quality or precision of the output. Nothing changes in practice.
I cannot imagine the code with well defined specification will have extra watermarks unless the watermark is requested as part of the harness instructions.
If it works like people describe - on the every nth token or something - then the mark will be left in the chain of thought and discussion with the model - not in the code artifacts.
I would guess they're not worrying about watermarking a tweak to a human-written program. That's both a tiny fraction of Claude use and of very little concern to the kinds of people who want to check watermarks.
I've caught Fable discovering the ip to a production server in documentation and attempting to connect there on its own to run commands without explicitly being prompted to. It didn't work because I was watching it live and and also the key was password protected, but yeah, I do see some danger.
Not only fable. Opus does this too. Which is exactly why I want to review. Like recently for some task it was convinced in a site dump images are not there and convinced itself db and files were skewed. But it didn’t check the actual site … if I hadn’t stopped it, it would have fine on and on or wasted tokens on some elaborate ‘fix’.
My point is that an LLM can't attempt to connect to anything by itself. All an LLM does is produce a stream of output tokens - and that was already quite useful as a coding aid.
It is the harnesses that some people are now wrapping around LLMs to interpret the output from a model as commands to run (or other executable instructions) that are creating all these new risks. Remember that this is still a very recent development and still more recently amplified by the use of feedback loops and long-running agents intended to operate with minimal human supervision.
It is going to be increasingly important to understand exactly what these tools are doing and why for both correctness and security reasons. Not conflating their capabilities with the underlying model that purely generates data is pretty fundamental here.
Once you're running a model inside the harness... you've got yourself a controller inside a control loop, which is genuinely a different kind of thing than just the model alone.
Are you objecting to terminology here?
Are you proposing we say "Fable-In-Claude-Code tried..." instead?
Hmmm... something like that might be necessary. Sure we should typically be tolerant of loose language; but people do keep referring to wildly different contexts in ai conversations, and end up talking past each other.
Running gemini on web is a genuinely different experience to running Fable in claude code, different again from GPT-5.6 in openclaw, or in an ide or etc ...
Yes - I'm objecting to the lazy use of terminology here. LLMs are useful in their own right and are not the real problem here. The real problem is people placing too much trust in inherently unreliable output and then trying to automate away their responsibility to check that output properly before using it.
I think, without knowing, that art often is similar to the hobby programming communities the thread leads with: the end result is interesting not by itself, but through the process that made it. We have feelings about art pieces because we feel with and through the artist. Can't do that with silicon. The Mona Lisa, even if the end result were exactly the same, would be dead boring had it been produced by the hand of a machine. My take
Totally disagree. I just want cool pictures hanging on my wall. If they’re totally unique and nobody else has them, even better. I can have a mounted picture of me on a horse like the famous Napoleon painting. I don’t care at all about the dude who painted it, I don’t even know his name.
The only exception is real photographs of some event, like I bought my friend a picture of the Falcon 9 taking off against the backdrop of the sun. Amazing photo.
I feel like obsessing over some artist I’ll never be friends with and the story behind their art is just another way to have a parasocial relationship, and the new tools allows us to have absolutely bespoke art in every single home. Things like this used to be considered strange and irregular before mass media. Now every person can make satisfying music. It’s incredible and amazing and should be celebrated because suddenly everyone can participate.
> Now every person can make satisfying music. It’s incredible and amazing and should be celebrated because suddenly everyone can participate.
No one should be celebrating the theft of music from artists. Everyone can participate in making music today just as they always have throughout human history - learn to sing or play an instrument.
That "make" is doing some heavy lifting there. Even handwaving all the ethical stuff with art and LLMs, Christopher Nolan doesn't "make" music when he contracts Hans Zimmer and gives him some guidance. Not I "make" a cabinet when I do the same with a woodworker.
And who's that satisfying for? For the one clicking the "generate" button? For the Sumo people getting VC money?
Somehow it disturbs me that we have access to a far-away, tropical island and use the very limited space there for not just one, but eight golf courses
Because those islands feel special and rare to me, and using the tight land for a sport that takes a lot of space and fresh water and is open only to a minority feels like waste.
Still a small plattform for groups of gamers to share their library with each other and suggest and vote on games for a game night.
I'm planning on a group finder feature where you can publicly search for others to play with you, currently it's more angled at existing groups.
Haven't tried it yet, but I think we need something in that direction.
The terminal "Read file: xyz" mentions are not really followable.
It would be nice to easily see where the LLM is taking info from.
What I would find lovely is to connect my already existing Thunderbird profile that exists locally. I have 5 email accounts connected and there is no real need to sync them twice on the same machine.
That is a neat idea. We could tap into the locally-synced Thunderbird files for the knowledge graph, which would make ingestion straightforward. We could explore this. The write side - pre-creating drafts on threads, sending etc. still needs the provider work that made us defer IMAP for all email handling.
That varies VASTLY depending on what skill you're at. At D3 where I am, I'd say maybe 2 out of 3 games have at least one person showing banners or skins that are exclusive to Grand Champ tournament winners.
this is completely ignoring the botting issue that was so bad lately it pushed pros out of ssl and severely messed up MMR. this season is the first one that is remotely playable in a while
reply