martialg

Author here - thanks for reading and thoughtfully replying

I’d personally love a world where open weights compete with proprietary ones, but I don’t believe it solves the core concentration issue. In that scenario most value still flows to capital holders, it’s just hardware holders not model weight holders.

I emphatically do not want to put the genie back in the bottle nor do I believe it’s possible. Technology has never been restrained for long (export controls on cryptography textbooks in the 90’s comes to mind here)

I also have already personally benefited a great deal from LLMs. I actually frame the entire essay series from this perspective in my prelude essay here: https://www.wysr.xyz/p/a-consigliere-on-every-desk-and-in

However, I believe we may disagree on the definition of a public good. If you’re referring to the free tiers of private models, then I’d argue that unless there is some legal framework passed that forces the frontier labs to offer that to everybody, it’s a customer acquisition cost laundered as a public good. It could disappear at any time and probably will as cutting edge model margins are reduced via competition.

In general, I believe the best AI policy balances allowing for maximum competitive market dynamics while hedging existential economic disruption risk for the general population.

I’ll go into this deeper over the next few essays. Appreciate the feedback

Author here! It's from a workshop in 2001 for the National Research Council's Board on Science, Technology, and Economic Policy. He gave a talk.

You can ctrl+f for it at this link

https://www.ncbi.nlm.nih.gov/books/NBK208682/

Glad you noticed. My personal preference is a more aggressive pursuit of a true public data commons, but for better or worse I am trying to write the essay series by balancing what I can be implemented sooner/with less pain + heartache and I think that discussion is a much longer debate with much harder policy to define.

Author here. This is a great point! I explored this case extensively and ultimately didn't include it in this essay because I didn't want to get bogged down in east coast riparian rights vs the west coast judgments. It is certainly a tumultuous time for contested public rights.

Author here. Thanks for reading and the kind words. I will talk about distillation and OS in coming essays (the is a multi-part series).

Author here. Appreciate your thoughts and I mostly agree actually.

I'll explain more over the next few essays, but I am designing my proposed regulatory structures to try to accomplish 2 purposes in tension simultaneously like the Fed: 1. Maintain global competitiveness for frontier labs 2. Create a societal hedge against the AI bull case (AKA the economic black hole case)

A % of revenue scales in a way that I think balances the two well while avoiding all the other problems I mentioned in the essay. I’ll get into ratchets, timing, and thresholds in later essays, but I agree the China/competitiveness problem is central.

Author here. Thanks for reading. I agree there are a lot of levers we can pull to work towards something better. I've structured this essay series as a sequence of nested regulatory solutions so in the next few essays I propose additional structures with instruments like this. They're sequenced in a way I believe that can be pragmatically implemented and start showing progress in the next decade or so by the US. So stay tuned!

Thank you kindly

I’m working through thoughts on this as well and agree with your read on the incentives.

There is an interesting set of conditions that happens if/when models get so competent that they’re effectively indistinguishable from each other and inference becomes a true commodity. IRL impact will lag this ofc but it’s such a wild time to be alive.

Author here. Really appreciate you taking the time to read and for the kind comment.

I think the tension between these ethical questions and the practical realities (both the good and the bad) of AI is likely the defining issues for technology and perhaps society in this decade.

It’s important we’re thorough and rigorous with how we think and act here so I really appreciate you engaging with the topic.

Author here. Thanks for taking the time to read.

I agree we’re in an interesting era where frontier research has shifted from mostly publicly funded to mostly private and it creates challenging incentive structures especially regarding externalized costs of research.

Did you have any thoughts on my argument of how public knowledge does get damaged by the proliferation of AI over time?

Author here. Thanks for reading!

I have additional essays coming out that will address this exact issue and other issues I know that people will raise.

I’m building the essays series around arguing for practical policy I believe can get implemented and am sequencing it as thoughtfully as I can. I just can’t fit every argument into every essay.

Eucrisa - https://www.goodrx.com/eucrisa?label_override=eucrisa&form=t...

fascinating!

As a veteran I think it's incredibly important for more people to see this.

Interesting that it specifically mentions “non-bargaining” fires only.

Super interesting

10x engineer is a good meme, but a bad classifier

I lift.

It’s an infinite series of stacked physical exertions that reminds me that I can’t think my way out of putting in the work.

It’s a great metaphor for how most really incredible projects are actually built (with a lot of grind). It also helps build the mental muscle of ok I gotta hit my targets today so I can hit my new targets tomorrow

I wrote a long form essay about the concept here if you’re interested.

https://www.wysr.xyz/p/why-clever-people-should-lift?s=w

30% of all cards were printed in the past two years.

Wow I’m so off on predicting trends it’s crazy

Also IMPORTANT: Don’t hungry shop!!

1. Make your grocery list. 2. Eat a snack. 3. Go to the store and shop. 4. Profit.

You will buy food that will take you further away from your goals if you hungry shop.

So fwiw I consider myself an athlete (ex-military, powerlifter, been lifting for 15 years) and I still struggle with nailing my weekly meal plan. It’s a nontrivial effort and organization problem.

There’s also no “best” way to eat without identifying what the metrics you’re optimizing for.

For me, I’ll oscillate between really motivated to be lower body fat vs not and that’s basically a function of roughly hitting macros (40% protein, 40% carbs, 20% fats) while staying below calorie counts (2500 on non lifting days, 3000 on lifting days).

It’s not exactly “right” but I can generally manage to hit it with 3 meals a day plus some snacks. If I can manage this without going out to eat during the week then I generally lose body fat over time because restaurant meals have a lot more calories and sodium than home prepped food.

The meal prep is more complicated on “high motivation weeks” but I work to maintain this as a baseline. Trader Joes is my grocery store btw.

Breakfast every day (pre lifting meal) 5 minutes of prep - Egg whites from a box with some shredded cheese + granola bar or handful of granola

Lunch every day: 5 minute prep - bread + Turkey slices + pepper Jack cheese + mustard and 3-4 rice cakes + seltzer water.

Snacks: Turkey Jerky or Yogurt

Dinner: 5 minute prep (I cheat slightly here) - Trader Joe’s prepared meals like a burrito or a Tzatziki wrap or a Caesar salad or a half a box of their chicken nuggets. I sometimes cook my own food but this is a good compromise between a fairly spartan homemade meal but not 1000 calorie restaurant meal.

Portions tend to matter more than you expect.

When in doubt, have a fixed rule like no carbs after 1pm and give yourself a free meal or day once a week for your sanity. It matters more for your life that you just start trying to eat better than to actually nail it off the bat.

Best of luck and feel free to reach out to me if you want to chat further! You’ll figure it out.

This is somewhat fair.

I guess my speculation stems from a lack of substance to poke holes in?

Even going line by line beginning with the first real paragraph we find they launch into a classic straw man of rejecting the "industry claims that crypto-assets an innovative technology that is unreservedly good". Nobody with a whiff of tech acumen in crypto thinks or says this. Additionally, anyone not thoroughly anarcho-Libertarian also gets the need for guardrails to squash scams.

A scattershot of the other tangible claims in the paper feel pretty similar.

"Append-only digital ledgers are not a new innovation. They have been known and used since 1980 for rather limited functions." <- true, but the also can be said for the evolution of AI models and we see where we're finally at with those today.

"Blockchain technology cannot, and will not, have transaction reversal mechanisms because they are antithetical to its base design." <- quasi true depending on the chain, but also we can approximate reversibility with clever escrow-like transactions (I'm actually building this full disclosure)

"Financial technologies that serve the public must always have mechanisms for fraud mitigation and allow a human-in-the-loop to reverse transactions; blockchain permits neither." <- unequivocally stating blockchain tech can't mitigate fraud in ANY way is silly and reductionist. Deciding human-in-the-loop reversibility is the pinnacle of fraud mitigation is also pretty wild considered how much effort and resources developers just on HN put into removing humans from their process execution loops.

"Finally, blockchain technologies facilitate few, if any, real-economy uses." <- partially true in the present tense (similar to how everything has to be adopted before it reaches mass adoption), but glaringly narrow minded if considering potential uses. Anywhere there is asymmetric trust and power between transacting parties one can imagine the possibility of an application! Not to say its necessarily the best use case, but the lack of effort of consideration seems intentional.

There's more, but I hope this provides some context to my initial post

VF Protocol | Cofounding CTO or Founding Engineer | Full-time | Prefer IRL, but Remote works | Boston, USA

VF Protocol is building a zero fee payment network on DeFi rails.

We're starting with P2P NFT transactions and iterating from there.

I'm too excited to wait so I'm building out the MVP myself (Solidity, React/Antd/JS, Alchemy, The Graph, and whatever else I need to learn next), but looking for a cofounder to scale the MVP into what the technology should become over the next ten years.

Website: https://www.vfprotocol.com/

Direct Contact: Cameron@vfprotocol.com

Diligence on me: https://linktr.ee/camarmstrong

So - I love thoughtful criticism.

My ego doesn’t, but I generally find it very valuable.

This letter is honestly not thoughtful criticism.

It reads like a group of academics and specialists annoyed they were too early to profit from a technology they spent a lot of time thinking about esoterically and avoids specificity in…well… any meaningful way.

It also makes no effort to segment the space.

If a critical letter created by supposedly credible experts in a field can’t spare a paragraph to delineate between DeFi, NFTs, L1’s, smart contracts, stablecoins, and whatever else is bothering them then I think it’s fair to assume they don’t know the difference.

And if they don’t know the difference then why are we even reading this letter?

It’s a waste of our time (and theirs) to try and unpack what is pretty clearly frustration at the fact that the effective application of basic research is all that matters for adoption.

I think (at its core) everything crypto is a novel economic incentive layer wrapped in a technology package.

This creates the opportunity for blazingly fast user feedback loops and incentive alignment that wasn’t possible before (for better or worse).

Taken from that perspective, this letter misses basically everything interesting about the potential of Crypto.

So.

Read it. Form your own opinion. And move on with your day.

Gm everyone

Cool - duly noted. Thanks!

Amazing - thank you

What do you mean by this?