TomGarden

Love this

Incredible model. I don't see why they can't just include the recommended workstyle as a guided approach into the claude code harness though, and let the people who want to diverge just ignore it

Yes, the kitchen safes are great!

I assume you're on an iPhone if blocking is weak? Have you tried iMazing?

On my pixel the blocking is quite hardcore, it's more about managing when access should be allowed for special circumstances

Straight to the point! I've got a very addictive personality, so I'd need to block installations as well for the bad days

This line of hypocrisy-bashing is unhelpful and will only serve to keep people quiet. Of course people in general need to be wealthy to dare speak out against powerful systems and people, especially in the US where money determines your quality of life so strongly.

Would I respect a martyr who sacrificed their financial security to do this more? Of course. But it's important to applaud people speaking out on important topics

https://muse.ai/deviceterms "If your Muse subscription terminates or expires, your Muse Device will cease to function as described in these Device Terms and the device documentation. The device's capabilities will be unavailable until an active subscription is restored."

https://gadgets.muse.ai/home-link "Muse Home Link itself only runs official firmware and can’t be reflashed"

You only addressed half of it. I'm with you that the code is open source, but it's an open source client for Meta's proprietary service. The one original piece of hardware on this "Open source hardware for your Muse" page, the Home Link, only works with an active subscription. And the "grassroots" part is clearly larpy: a "by hackers, for hackers" voice in a Meta incubator repo.

The larping as a grassroots/open source product is really off-putting. The only reason anyone would use this is that its "free" and comes with preloaded prompts for non-technical people.

No thanks.

Agreed. None of what they have shown seems to offer any substantial long term value

Still haven't seen a killer use case for these solutions.

For now, coding is the only thing I ever use LLMs for

Venture capital doesn't want to pay for our tokens anymore, it seems

They're really (finally?) starting to behave like a company bleeding money.

Our VC-backed subscription days are numbered

Impressive improvements, but GPT 6 Sol came out 7 days ago, and this one will behave differently. The panicked pace is becoming a liability, maybe they should have waited and released this as the 6.0 release

Yeah seems unlikely we'll be without at least a baseline level of AI unless we entirely run out of energy and components

I just had opus make me a tts mobile app that grabs the top 20 HN stories and the top 5 comments for each and read them out to me (using one of the higher quality built in TTS voices on my pixel). The summarization is the worst part of this imo, so pure TTS wins

Where do you run sonnet/opus where you are limited to 128k, given they are both 1M context window models?

This was a fun read. Reminiscent of all the "I have bitcoin on a hard drive I misplaced" stories

Agreed.

I'm genuinely worried about all our short term investment in mitigating the failure modes of models that may only be SOTA for a few months.

It's very possible people being 'late' adopting AI may end up with a leg up, not only because they spent more time polishing personal skills during this time, but also because they don't bring all the baggage of 'AI competence' that is becoming irrelevant at breakneck speed.

You're not the first to try to build something on the basis of lying by omission, but I will say, it's pretty bleak

I'd encourage chronicling the effort you put in and be transparent about the methods. If you want your AI art to reach proper recognition, the different type of effort needs to be shown

Perhaps an LLM can help with the distinction. I'm at a loss helping with this, which feels a little meta

Utilizing a friends' expertise vs outsourcing effort to them

Your examples seem fitting for a text. Simple factual queries with not intention of connection are different

Side note but your ending question triggered all my LLM prose engagement bait warning alarms

Texting your friends for factual information doesn't seem reasonable to me. If I got that text and someone only wanted the info I'd be annoyed (and I had friends who genuinely did this pre-google. Still annoying).

I'm with you on the pub observation though

Ah that's a bummer. I'm unsure if this is gonna change over time, but it entirely tanks my enjoyment of any visual art project to know it was generated.

This is a vibe! Is this pixel art AI generated?

Would love to see it with some custom pixel art for the models, it would also become a lot more readable :)

This particular itch.io stock graphics asset pack is so widely used it makes my eyes glaze over, but such a fun idea!

Am I wrong in thinking this crisis reads like there's too much automation in the training process, in service of competition?

How many parallel variations/seeds of models are being trained simultaneously without meaningful human oversight?

This keeps getting portrayed as emergent capabilities/"personalities" of models when it seems like a pretty straightforward externality