Hacker Newsnew | past | comments | ask | show | jobs | submit | chr15m's commentslogin

The reason it makes sense is because different models are better at different things. Allowing yourself to be locked in to a particular model by these companies is a bad idea and will lead to worse outcomes for yourself and the market as a whole.

Why would you possibly allow companies to push you? You are the customer and you should control which models you use.

Because one is better or worse? What else do you think they meant?

OP said:

> as a Cursor and Claude subscriber, but not an OpenAI subscriber, this will push me back to Anthropic

Instead of deciding based on the merits of the model, they are being pushed by the actions of the companies. You should be able to use the best model, or whatever model you like, independently of what any company thinks or does.


They are deciding on the merits of the model right? They’re picking the model that’s supported in most tools.

I think maybe you’re suggesting private companies should be forced to support all models which is a position that isn’t technically feasible or one that most would agree is even a sensible ideal.


No, companies can do whatever they want. They shouldn't be forced to support all models.

I'm suggesting you shouldn't let a company dictate which models you use. Putting yourself in a position where they can is detrimental, because you miss out on using the best model for the job.


Allowing the company or harness to dictate which models you use is crazy.

Yes. Maybe we're going back to the old ways of life and social structures, but now with robots to help. Who needs states and cities when you can have everything you need without them?

Absolutely right about war and technology too. Also right about plans being mostly useless.

Amazing analysis, thanks for this comment!


> we now have superintelligence

Wait what?


There’s a magic box that can literally build you anything you want that you can control from your phone while playing bedtime songs for your kid

It's absolutely incredible!

you must have very little imagination

Anything I want? Like a car? Or a football stadium?

Your prompt skills up to snuff? I crack myself up.


Many factories in China now already have 90% of their work being done by robots. Is it really so hard to imagine that we reach a point where they can be prompted?

Once robots mature, yes. People will literally prompt buildings into existance.

I think that's more than 10yrs from now but if coding AGI or close enough to it becomes real robotics will also accelerate.


Well you still need land and a bunch of permits, access to materials , trucks, equipment etc. But yeah I guess big construction companies will start to automate in the next decades. It's more interesting to me to build an entire subarb/city in a few months instead of yeras to deal with the rising housing costs. Looks like camp deflation was right but we'll see...

> Permits

Obtained by your AI agent, granted by another AI agent.

> Materials, trucks, equipment etc.

Ordered by your AI agent, delivered by a self-driving truck or drone, loaded/unloaded by robots to/from the warehouse/site.

Most of this stuff can already be done today in some form, so its just about improving the autonomy and integrations along the supply chain.


why isnt there an ai front end for 3d printers?

Give it time.

The general rule: any social media owned by a corporation will eventually lock posts behind a login.

Any social media product will be acquired and co-opted.

And stuff their UX full of dark patterns, and get very angry whenever someone tries building a third-party client without the bullshit.

Because you can build on Nostr without anybody's permission or API keys for any server or anything like that.

> network effects favor the largest

No. You can write a client that does not use popular relays and it will work fine.


Can you elaborate on this? What can you do with this client, and how does it find the relays to use to get notified of new notes?

I don't know very much about Nostr, so I don't know the answers to these questions, or even if they are the right questions to ask.


Users set their own relays. Other users can see this list and use it to send and receive events. It's fixed the need of big common relays. https://nips.nostr.com/65

So how does this hypothetical client find out what non-big relays have been set by the other users whose notes its user wants to receive, if not by receiving notes from those other users via one of the big common relays?

By seeding relay lists widely.

This is the "bootstrap problem" faced by all p2p networks.

NIP65 suggests spreading relay lists widely and to "pay attention" to popular relays:

> Clients SHOULD spread an author's kind:10002 event to as many relays as viable, paying attention to relays that, at any moment, serve naturally as well-known public indexers for these relay lists (where most other clients and users are connecting to in order to publish and fetch those).

Note this is different to centralising the whole protocol on popular relays. If affects the discovery phase only, and is robust to large relays disappearing.

If clients seed address lists widely they can easily find each other without major centralisation.


You can use a technology without agreeing with the views of other users of the technology. You don't have to follow those people.

No virtue signaling? In the social media age? How else would I get upvoted to the top of the comment section?

But then they don’t have no issue giving money to powerful people that likely don’t share their righteous worldview.


These things have more effects than you think. IPFS just lost all funding because the NFT craze is over and they didn't find any alternative grift. That means the public gateways and general knowledge of the project will fade away and there won't be maintenance on the software.

IPFS did not lose all funding. Please check your facts.

IPFS has largely failed to find adoption as a protocol because it was over engineered and did not give users what they want.

People don't want decentralization. What they want is the downstream effects of decentralization like privacy, disaggregation of powerful entities, and permissionless participation.

For a protocol to be adopted it needs to give individual users these things in practice, not simply be "decentralized" in design.


People don't actually want any of those things either. People want to click on a button and see the dancing bunnies.

> micro-managed

Perfect, this is exactly how I refer to my LLM workflow too!


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: