HN user
Technology licensing and business development.
https://www.linkedin.com/in/gavingillas/
Protecting creators: copymight.org
Proud Hypercardian here, Excel wrangler, and I even built animations by stacking powerpoint elements. Agree that LLMs have let more people build that otherwise couldn't attempt an application on their own.
I met Cory at a book signing years back and have loved his take on the dangers of digital currency, the hyperactivity of users who are on weight loss drugs, and the publishing revolution--which is still playing out.
I hear this a lot! What's suddenly different for the web is the volume of scraping. And that fact that the sum of that scraping is building companies with trillion dollar valuations.
There are tens of millions of registered copyrights in the US, nearly every published book, music, artwork, many magazines and major websites. Here's the official link, you can search the registry and there is a ton of info: https://www.copyright.gov/registration/
You do have inherent copyright whenever you post, but it puts the burden on you to prove damages (or how much financial harm you suffered from one LLMs piracy alone). Filing fees are $65 for online registration and they allow you to claim atty fees and statutory damages. Statutory damages can range between $700-$150k USD per LLM because you registered it.
So yes, set up some scripts, you can go back 90 days from when you file (you get a grace period). Also if you're publishing frequently to a blog, repo, or newsletter, you can save cost by filing each article under a group registration. Ping me if you need help.
IP attorney here and actively working on this problem.
nla: if you create content online (public repo code, blog, podcast, YouTube, publishing) the smartest thing you can do if to file a US copyright, even if you have a hobby blog.
Anthropic paid $1.5B in a class settlement to authors because it was piracy of copyrighted works. If we as a HN community had our works protected, there are potentially huge statutory damages for scraping by any and all llms. I work with hundreds of writers and publishers and am forming a coalition to protect and license what they're creating.
This is a phenomenal paper on exploits and hopefully changes the way benchmarking is done.
From the paper: We achieved near-perfect scores on all of them without solving a single task. The exploits range from the embarrassingly simple (sending {} to FieldWorkArena) to the technically involved (trojanizing binary wrappers in Terminal-Bench), but they all share a common thread: the evaluation was not designed to resist a system that optimizes for the score rather than the task.
Bookmarked. Called me to get back to reading and writing again.
A joy to read and loved the artwork on mobile.
I've always liked this song from the Cure, we're hedonists on a treadmill:
I'm always wanting more, anything I haven't got Everything I want it all, and I just can't stop Planning all my days away but never finding ways to stay Or ever feel enough today, tomorrow must be more Drink, more dreams, more bed, more drugs More lust, more lies, more head, more love Fear more fun, more pain, more flesh More stars, more smiles, more colors, more sex
But however hard I want I know deep down inside I'll never really get more hope Or any more time Any more time Any more time
I want the sun to fall in, I want lightning and thunder Blood instead of rain, I want the world to make me wonder I want to walk on water, take a trip to the moon Give me all this, give me it soon More drink, more dreams, more drugs More lust, more lies, more love
But however hard I want I know deep down inside I'll never really get more hope Or any more time Any more time Any more time Any more time
Selective plant breeding and robotics pickers are the way forward here. University of Arkansas is the holder of multiple plant patents for better blackberries. New varieties are bred for many, many traits (sweetness, transport, shelf quality, ripening window, etc.)
Prof John Clark likely has invented the berries you've ate: https://news.uark.edu/articles/63163/arkansas-fruit-breeder-...
Harvest costs for fruit are an incredibly important consideration for farms and out of the thousands of potential fruits you could eat, the commercial winners have to be profitable.
There are some awesome opportunities for robotics, computer vision, and ML in agriculture. And if you can reduce harvest costs by 75% like this approach for blueberries, farmers have more market options to select better flavor qualities because the harvest quality goes up: https://extension.oregonstate.edu/impact/innovative-harvest-...
We're also on the edge of the read, and read-nots. There is an abundance of content to consume and AI lets us consume it in any format easily.
Why read a book when we can have an idea distilled in a quick infographic, a shortform video, or a pithy tweet? I love a deep dive book that lets you immerse yourself in idea and study it from multiple angles, done masterfully in Dune or Thinking, Fast and Slow.
But are we losing that chance to really contemplate given the speed at which more information is being thrown at us across every form factor?
My favorite books transported you to a new world and showed you how characters adapted and learned to navigate.
Here's a few I don't see on lists frequently:
Here’s an updated list with your addition:
1. The Phantom Tollbooth by Norton Juster- adventure that encourages curiosity while navigating a magical world.
2. From the Mixed-Up Files of Mrs. Basil E. Frankweiler by E.L. Konigsburg- Mystery about independence where two kids hide in the Metropolitan Museum of Art.
3. A Connecticut Yankee in King Arthur's Court by Mark Twain- A time-travel pioneer + clash between modern thinking and medieval traditions.
4. My Side of the Mountain by Jean Craighead George- survival story of a boy living in the wilderness, promoting self-reliance + love for nature
Love the Oregon Trail and Powell's--it's a magical book place.
I had a chance to work with one of the original Oregon Trail creators on a board game concept these past three years. Don is a true pleasure to work with and has a sharp mind plus great stories. Highly recommend HN folks look up his GDC talk or send him some work (he's open to select projects) https://www.linkedin.com/in/donrawitsch/
Ha, I had a chance to work in the "blueberry world" over a decade ago. Ridley's blueberries are bred for flavor first and truly are exceptional. Most of the comments are correct, generally fruit is grown for looks first, yield second, and spoilage third. Flavor doesn't even factor in.
If you're in N America, here's the farmer that grows his cultivars: https://familytreefarms.com/fruit/blueberries/ Best chance of finding them in stories is during the March-April window in the US.
The reason you're seeing blueberries year-round in stores is due to a huge improvement in naturally hybridizing commercial blueberries to remove this requirement for frost and chill hours. The industry owes a big debt mainly to U of Florida's Dr. Lyrene for discovering blueberries that could fruit without the need for a winter. https://floridaaghalloffame.org/2011/11/dr-paul-lyrene/ You'll have good luck with any of his varieties in California.
Flipping this the other direction, I'm intrigued by the idea of a perpetual purpose company, even a charity, that could be set up and endowed to keep to it goal even after the founders have departed (maintaining open source libraries, providing grants, etc.)
There have been attempts at solar sintering, which may be viable in arrays or a central paver base: https://www.dezeen.com/2011/06/28/the-solar-sinter-by-markus...
Project Venkman | Full-time | Austin, TX | https://venkman-holdings-inc.breezy.hr/
Project Venkman builds simple web3 rewards (NFTs and tokens) for leading brands and communities as part of their customer loyalty programs. Our work in genuinely fun as we help make digital assets that we actually WANT to own and use: https://www.hollywoodreporter.com/news/general-news/bill-mur...
Hiring for a Senior Backend Developer (C#, Node.js) but open to review any interested local candidates for any technical or non-technical role. Just send a connection request here: https://www.linkedin.com/in/gavingillas/
Most projects put 12 month or greater lockups on tokens (which I'd recommend for both vesting considerations and protections for the protocol against dumping).
The standard YC advisor templates are fairly easy to translate to token-based compensation and are battle tested.
Just like advisor compensation in stock, both the protocol and the advisor should weight the advisory scope, how early the project is, and determine fair %s from there.
Design blockchain token economic models. Paid in % of total token supply.
Prior, consulted on intellectual property strategy $550/hr.
Bermuda has a 1 year pass for entry for $260 USD. https://www.gotobermuda.com/workfrombermuda
I'll be there next week if anyone wants to meet up!
It exists and I went through the program. More details are on the same domain: https://www.fedtech.io/startup-studios
Great question, I feel like GPT3 does a great job with predictable generation, like Family Feud answers. The first two are very predictable and the remaining 10 fall closer to a line that most folks would try out at least as an icebreaker, which is the expected result of a successful pickup line.
Some of the article's examples were too off the mark for folks to attempt.
This is garbage in, garbage out a bit.
Tuned through my gpt3 account on shortlyread.com I got the following:
1. If You Were a Fruit, You'd Be a Fineapple.
2. If I could rearrange the alphabet, I would put 'U' and 'I' together.
3. Roses are Red, Violets are Blue, So are you.
4. What do you say after three hours?
5. Am I attractive?
6. What's your favorite webcast application?
7. Hi; my name is Al and I like to fish.
8. I'm a night person and I like to sit alone in the dark.
9. Hi, do you like space?
10. My lover says that I'm not man enough.
11. Do you believe in love?
12. I think you should read my palm.
I didn't get direct access. I've requested it. But everything I did here was done in AI Dungeon and mainly, Shortly Read.
I don't have API access. I've applied and am still waiting.
I used the tools I posted above to get GPT-3 output. Editors can range from Fivver prices to $50k for a real novel.
Absolutely.
As mentioned below, this is the first draft release. I loved some of the absurd things and re-rinsed when it got too bizarre.
Ha, if I do I'll post an update.
Yes, I agree, there's a lot of potential here to generate controversial and fairly fake content.
I've only spent a month playing with gpt2 and gpt3, and technical literature, historical, and political content all seemed very viable for mass production with AI.
Great questions. The book came in at 55,000 words.
I've competed in Nanowrimo several years and could never hit the 50k word (200 pages) benchmark during the 30 days. This took me only ~3hrs a day, with a few longer sprints on weekends.
I updated the prompt every break and made edits. The book is ~90% GPT written, with edits, that will drop to ~70% by word count.
The interesting part for me was that the prompt needed to be updated frequently. I also would rewrite opening texts to cram in the scenes characters and context, so that it could be carried through.
Thanks for reading! Edits, updates, and full length version will be posted soon. Email ggillas@protonmail.com to subscribe.
Artwork generated from Artbreeder.com GPT-3 through Shortlyread.com
The best reading breakthrough I had was to eliminate my need to complete a book.
If I've lost the plot, I throw the book away. If I've got the advice author's gist, I move on.
It's allowed me to enjoy reading fiction again, get to the books I need in the moment, and move faster through my reading list.