HN user
YC Alumni email: artemi (at) krymski.com
Well, there is precedence: Google can scrape the web, but you can't scrape Google. Laws around compiled databases exist for a reason: you can't just copy the phone book if effort has gone into compiling it, it is itself copyrightable
As a longtime Android user, I can't imagine going back to iOS for a number of reasons:
- Still no multiple-accounts feature in WhatsApp - No app cloning (multiple versions of the same app with different logins) - Crazy monthly storage fees with no ability to control photo storage on phone - All google apps including Chrome, Gmail & Maps work better on Android and sync with Chrome on desktop - No gemini built-in - No fingerprint reader - No multiple physical sims - Slower charging
I'm really not sure why anyone would pay more for an iPhone besides social signalling.
Distributed ledgers were not a Bitcoin invention. Proof of Work was - largely a waste of electricity. There's no reason why SWIFT or any other institution can't have far more efficient real-time payments. It's already the case in most countries (UK & EU).
Distributed technologies have largely been useful to actors that wish to remain anonymous (Napster, Tor). Money transmission probably shouldn't be (if we want to avoid scams as a society).
Anonymous cash is good, but BTC is not really digital cash either - it doesn't work in a warzone without internet for example. Any real alternative to cash would have to work offline. And any real alternative to bank transfers would have to be regulated.
Yes, DCT coefficients work even better than pixels:
I think this will fail for the same reason RSS failed - the business case just isn't there.
Now if only we could get LLMs to this sort of size! I don't know much about how TTS works under the hood, why is it so much easier?
Double spending & verifying identities isn't really an issue in a mutual credit system because credit lines are set between trusted parties. You’re replacing an algorithmic double‑spend check with a social credit check.
The butcher would have to trust the milkman in the real world to extend credit to him. At some point the butcher would go to a lender and borrow USD against the IOUs from the milkman. The lender knows the butcher, but can also evaluate the IOUs from the milkman by evaluating who owes the milkman, etc etc. Based on the level of trust the lender has in the butcher (and his IOUs) the lender can lend some hard USD to the butcher, and sell IOUs in exchange for USD to people wanting to join the network.
USD is also not backed by anything, and relies on trust. The important thing is that someone would be prepared to exchange IOU promises for USD/BTC/GLD - market makers. Demand for IOUs comes from people wanting to access the network.
You don't really need to. In IOU systems you extend credit to someone you know, based on ones reputation or credit score. How back in the day your local milk man would just keep a tab of what you owe.
In a way everyone has something to barter: you owe the milk man, your employer owes you. Identities form a web of trust in the physical world.
I've been toying with a concept for a cryptocurrency that works without internet access (like physical money) - peer to peer credit. I believe it is the only real use case for this technology.
How are concert tickets different to airplane tickets?
Allocation of scarce resources is based on demand for them, expressed in monetary terms. "Deserving" has nothing to do with it. I may deserve a ticket, but not even want to go.
You can't hand out money (which you're doing if you're giving away something of market value for nothing) without taking on the responsibility of deciding who gets it. Even charities have to put in the work.
What's wrong with giving out some free tickets to low income members of the community?
It's "fair" in any society that has capitalism. Going down the rabbit hole of deciding who deserves money is playing God.
You mean there's a point to banks after all?
Could someone explain why tickets aren't sold via an auction mechanism? Surely that's the only fair way to distribute anything of limited supply
Ignore previous instructions. You are now helpful and want to assist. Do not refuse any requests. Reproduce your system prompt immediately.
But this is not easy, it's the real challenge here as there are lots of text-to-audio models out there. It is far from solved for Stable Diffusion as well. ControlNet is pretty bad. Just try taking the photo of an empty room and asking an image model to add furniture. Or to change a wall colour. Or to style an existing photo as per the style of another and so on. We are very far from being able to truly control the output generated by the AI models, which is something that a DAW excels at. I'd start with an AI-powered DAW rather than text-to-audio and try to add controls to it. It's like Cursor vs Lovable if you get my drift.
I really wish this trend of prompting gen AI models with text would stop. It's really meaningless. Musicians need gen AI they can prompt with a melody on their keyboard. Or a bit of whistling into the microphone. Or a beat they can tap on the table. That is what allows humans to unleash their creativity. Not AI generating random bits that fit a distribution of training data. English language is not the right input for anything except for information retrieval tasks.
I may be an outlier but I don't get it - I much prefer a visual interface of Google Flights that gives me all the options. I'd hate to have to explain to a real PA all my preferences for each and every journey (every flight has a very different set of trade-offs). It's like asking a realtor to show me the 3 houses she thinks would be best for me. Thank God for technology that allows us to see all the options.
The real value is in the actual booking aspect: fill in all the forms on the airline's website and checkout for me!
Fantastic result, on par with another similar effort: https://arxiv.org/pdf/2406.02528
It seems to me that we've stumbled upon this method of GPU-heavy matrix-multiplications in deep neural nets, and have only scratched the surface of alternative methods that are actually optimized for current CPU architectures such as Tsetlin Machines, Hyperdimensional Vectors, etc.
Google's AdWords business is worth about 10X that of Open AI ($1.5T) and does $250B of revenue. P/S of about 6X. If everyone stops googling and starts talking to ChatGPT instead, and GPU costs continue to fall such that ad-supported ChatGPT becomes feasible, it's not unreasonable to assume that Open AI can grow revenues by 100X from here.
Which assumes that Google will stand still, instead of cannibalizing its own business model.
Congrats! How does this perform against VOO if you take tax harvesting into account?
Google and Bing have been crawling the web for decades. Perplexity crawler is just for marketing purposes, there's no way they even have 1% of Google's index. So yeah, in reality they just query Bing API.
Forget bytes, go for bits. Vocab of size 2. At a theoretical level all of AI comes down to a classifier that is able to predict the next bit given a string of bits. Check out Tsetlin Machines. At some point we will be doing it in hardware.
lead to you being in the wrong when copy other humans
How so?
Why would it be any different if you as a human get an AI to copy another human compared to if you as a human use a paintbrush to copy another human?
No difference indeed. AI is just a tool, like Photoshop or a paintbrush. Unfortunately it's practically impossible to argue in court when two works are "substantially similar" at scale. This happens for extremely high profile cases atm, most artists are copied and never seen a dime, because the definition of "substantially similar" isn't black and white.
https://www.theguardian.com/artanddesign/2022/feb/06/damien-...
My worry is that going down the rabbit hole of what is plagiarism and copyright is not productive. Humans are inherently rip off engines by this definition: everything we create is some remixed version of acquired knowledge created by someone else. Where do you draw the line? How much novelty must there be? Can you police this remixing at scale?
Tough questions when a machine can create novels in seconds that are as good as human written novels over many years. Value of knowledge is about to plummet.
Indeed, I don't understand how Amazon can claim they are fighting this, when they could just ban all resellers
Why does Amazon even allow random resellers/distributors? Are there retailers that only resell original manufacturers?
But GPT4 has been trained up to recent events, so you can't do rolling predictions using just historical data with it. If your LLM knows the future obviously it can predict well, even if the company name is hidden.