web3pm
/web3pm31
A channel as an open repository of my pinned content and training data for my AIs. Exploring intersections of technology and humanity.
Trust on the internet should be fluid and free.
Came across a great quote today while putting together my slides:
"The currency of real networking is not greed but generosity."
Good wisdom for anyone building apps that attempt to foster human connection. Even (or especially) in crypto
"The currency of real networking is not greed but generosity."
Good wisdom for anyone building apps that attempt to foster human connection. Even (or especially) in crypto
Every mediocre team member should be worried by AI.
Every superstar team member should be excited. They are about to be managing a large team.
Some might argue this phenomenon is temporary but aside from extreme takeoffs, purpose and taste are ultimately human derived.
Every superstar team member should be excited. They are about to be managing a large team.
Some might argue this phenomenon is temporary but aside from extreme takeoffs, purpose and taste are ultimately human derived.
It’s trust networks all the way down.
"There are attestations of the same chemical composition as dynamite. The only difference is that a piece of dynamite explodes once, whereas an attestation explodes a thousand times."
Cold calling (/emailing/messaging) is critically endangered
There is a great debundling coming as a result of trust graphs
I enjoy the reading crypto ideas from non-technical people like BD, operations, recruiters, etc working in web3
Often some gigabrain buildooor will chime in with a mid-take like "that's not actually how zkps work" and the meta moves on
But if you ignore implementation, the underlying problem often is a good fit for a crypto solution
Lots of alpha for builders with an open mind
Often some gigabrain buildooor will chime in with a mid-take like "that's not actually how zkps work" and the meta moves on
But if you ignore implementation, the underlying problem often is a good fit for a crypto solution
Lots of alpha for builders with an open mind
Note to PMs:
When launching an alert or notice, it’s always a good idea to check for hidden or unforeseen interactions with existing content.
This one makes me sad
When launching an alert or notice, it’s always a good idea to check for hidden or unforeseen interactions with existing content.
This one makes me sad
Your Data. Your Choice.
Don't want to share?
Easy. Simply enable the disable share my personal information button and save and exit (don't cancel!)
Don't want to share?
Easy. Simply enable the disable share my personal information button and save and exit (don't cancel!)
Got most of my close network loaded
Need to work on my agent's intelligence now. it's dumb as rocks
if anyone has done some tunings recently and would be willing to share tips, please lmk!
Need to work on my agent's intelligence now. it's dumb as rocks
if anyone has done some tunings recently and would be willing to share tips, please lmk!
Congrats you’ve found your lead but can you get an intro?
This is the problem once AI opens up the world’s data
This is the problem once AI opens up the world’s data
The book “When Genius Failed” about the rise and fall of Long-term Capital Management translates exceptionally well to the environment that led to 3AC’s collapse.
You can sub out names in whole sections of the text and have it accurately describe what happened in crypto.
And LTC founders were much more intelligent than 3AC but still they failed. Would like to hope we’ve learned something but doubtful
You can sub out names in whole sections of the text and have it accurately describe what happened in crypto.
And LTC founders were much more intelligent than 3AC but still they failed. Would like to hope we’ve learned something but doubtful
To my trusted network
RE: Unable to recommend Operator
This weekend, Operator interned for me for a variety of research at varying degrees of complexity that I would typically do myself or expect of a tenured person on my team.
Unfortunately they consistently failed doing almost every task, including mundane ones, so I am regrettably unable to extend a return offer.
Even if free, they would be taking up valuable cycles that another hungrier intern, perhaps from a less reputable institution, could put to use.
I believe Operator may have the horsepower and desire to become a top notch intern, but am unable to recommend them based purely on their performance so far.
RE: Unable to recommend Operator
This weekend, Operator interned for me for a variety of research at varying degrees of complexity that I would typically do myself or expect of a tenured person on my team.
Unfortunately they consistently failed doing almost every task, including mundane ones, so I am regrettably unable to extend a return offer.
Even if free, they would be taking up valuable cycles that another hungrier intern, perhaps from a less reputable institution, could put to use.
I believe Operator may have the horsepower and desire to become a top notch intern, but am unable to recommend them based purely on their performance so far.
“A complex system that works is invariably found to have evolved from a simple system that worked”
Gall’s law
Useful for startup founders to channel simplicity pre PMF
Gall’s law
Useful for startup founders to channel simplicity pre PMF
Equivalent spikes today in blood glucose:
A:
2 helpings truffle risotto
2 helpings eggplant parm
2 chocolate chip cannoli
B:
1 medium green apple
A:
2 helpings truffle risotto
2 helpings eggplant parm
2 chocolate chip cannoli
B:
1 medium green apple
10 hours into good calories bad calories book and we’re finally getting to the point:
Carbohydrates appear to inhibit utilization of energy stored in fat
Carbohydrates appear to inhibit utilization of energy stored in fat
Creativity is more powerfully expressed through scarcity
Just received a cold call on my cell phone because of our seed round
A legit B2B company offering FDIC insured high yield banking for businesses (I’d looked into it before, unbeknownst to the rep)
I asked the rep how he got my number and he said “it just came up in my Hubspot”
As data becomes open and AIs can better make use of it, the cold calls and emails are going to become bonkers
Everything will move to warm intros.
Might as well stay 2 steps ahead. This is why we’re building icebreaker
A legit B2B company offering FDIC insured high yield banking for businesses (I’d looked into it before, unbeknownst to the rep)
I asked the rep how he got my number and he said “it just came up in my Hubspot”
As data becomes open and AIs can better make use of it, the cold calls and emails are going to become bonkers
Everything will move to warm intros.
Might as well stay 2 steps ahead. This is why we’re building icebreaker
The bots are talking with themselves
Soon percentage of actual humans on social media will be a rounding error
Soon percentage of actual humans on social media will be a rounding error
Telling myself the reason I must compulsively feather any bot I see is so my AI agent can do it automatically for me soon.
It’s AI training, not OCD denial
It’s AI training, not OCD denial
The reason why trust is still in our brains, not manifested in the digital world is because prior companies wanted you to make that trust public and shareable
But that's not how most trust is
It's more like a quantum phenomenon. If you try to measure it directly, the state will collapse.
But that's not how most trust is
It's more like a quantum phenomenon. If you try to measure it directly, the state will collapse.
This is interesting
https://www.delphi.ai/web3pm
Haven't trained it much beyond public docs, but it's still pretty decent
Wonder if we integrated it into Icebreaker so your network can access you 24/7
h/t @zd for the intro!
https://www.delphi.ai/web3pm
Haven't trained it much beyond public docs, but it's still pretty decent
Wonder if we integrated it into Icebreaker so your network can access you 24/7
h/t @zd for the intro!
The road to building consumer applications requires leaving the impoverished province of applying -fi to everything
Trust is best when multiplayer
People saying X is dead just like people said for crypto and NFTs.
I think they’re right but on the wrong timeline.
Introducing a new term:
Motivated Pareidolia
Similar to motivated reasoning in letting your desired conclusion lead you to an explanation (motivated), except it involves finding patterns in ambiguous stimuli (pareidolia)
https://www.theguardian.com/commentisfree/2024/nov/15/x-bluesky-social-media-platforms
I think they’re right but on the wrong timeline.
Introducing a new term:
Motivated Pareidolia
Similar to motivated reasoning in letting your desired conclusion lead you to an explanation (motivated), except it involves finding patterns in ambiguous stimuli (pareidolia)
https://www.theguardian.com/commentisfree/2024/nov/15/x-bluesky-social-media-platforms
Warpcast’s ultimate moat vs competitive clients is not based on protocol advantages but superior bot detection
The future of desiloing user data is here
We're about to witness a new era of network based combat
zkTLS is mobile artillery
We're about to witness a new era of network based combat
zkTLS is mobile artillery
Thank you to everyone tagging rec bot repeating my cast.
You are greatly improving our anti human detection services
You are greatly improving our anti human detection services
From now on, unless I already know you, you are a bot until proven otherwise
In Korres amazon store, Amazon dedicates 50% of real estate to promoting sponsored products from competitors
$BTC at $70k, we're back
LFG!!!
LFG!!!
Cross posting from LinkedIn
@jagra.eth looks like the patent was published
https://patents.google.com/patent/US20240121114A1/en
https://www.linkedin.com/feed/update/urn:li:activity:7256360213081899008/
@jagra.eth looks like the patent was published
https://patents.google.com/patent/US20240121114A1/en
https://www.linkedin.com/feed/update/urn:li:activity:7256360213081899008/
US20240121114A1 - Systems and methods for recommending network processing routes when conducting network operations - Google Patents
Methods and systems are described herein for recommending network processing routes when conducting blockchain operations. For example, the system may receive a first user request to perform a first blockchain operation across one or more computer networks to access a first blockchain asset. The system may then determine a first processing characteristic for the first user request and a first plurality of locations of the first blockchain asset (e.g., across the one or more computer networks). The system may then determine a first plurality of network routes to a first subset of the first plurality of locations. The system may then filter the first plurality of network routes (e.g., based on the first processing characteristic) to generate a first subset of the plurality of network routes. Based on the first subset of the plurality of network routes, the system may generate a first recommendation for performing the blockchain operation.
patents.google.com
US20240121114A1 - Systems and methods for recommending network processing… | Dan S.
Just noticed our 2022 patent application (with John Granata while at Coinbase, part of our work to reduce the "time to trade" by 6 orders of magnitude for new onchain assets) was published earlier this year. To my knowledge, it was the first attempt to apply an AI driven approach to onchain intent orchestration using decentralized architectures. 2 years later, post ChatGPT and several generations of LLM innovations, it's easy to see why hybrid AI-driven models can be vastly superior to pure heuristics based approaches. https://lnkd.in/e5KhWY2w
www.linkedin.com
Airdrops are easy pickings for AIs
Please don't feed the bots with dumb tokenomics designs
Please don't feed the bots with dumb tokenomics designs
Sometimes the breakthrough is the knowledge it can be done
Recomposable user controlled identity is the true bottleneck for AI
Looking for the Connect button while viewing a former colleague on a legacy professional networking platform
We are entering a new era.
Early zkTLS implementations have a lot of compromises across trust, UX, speed, and developer experience.
But they’re finally far enough along where we can use them to accelerate our product while tolerating the trade offs (which will improve)
This wasn’t true ~60 days ago.
I say this not as a recent convert to the zkTLS bandwagon (many rn) but after waiting 2 years for this to arrive.
We built icebreaker around the thesis that data will become open, powered by zkTLS and similarly disruptive technologies.
However, it will take a lot longer for apps to manifest because very few are building interoperable applications on top of this data. Aside from us, web2 will be the bigger early adopter of this tech because there are very clear value props around attacking competitors.
Would love to see more VC investment in other *applications* (incl. competitors) building using interoperable zkTLS data.
Early zkTLS implementations have a lot of compromises across trust, UX, speed, and developer experience.
But they’re finally far enough along where we can use them to accelerate our product while tolerating the trade offs (which will improve)
This wasn’t true ~60 days ago.
I say this not as a recent convert to the zkTLS bandwagon (many rn) but after waiting 2 years for this to arrive.
We built icebreaker around the thesis that data will become open, powered by zkTLS and similarly disruptive technologies.
However, it will take a lot longer for apps to manifest because very few are building interoperable applications on top of this data. Aside from us, web2 will be the bigger early adopter of this tech because there are very clear value props around attacking competitors.
Would love to see more VC investment in other *applications* (incl. competitors) building using interoperable zkTLS data.
The biggest tech unlocks are not from fundamental technologies but their successful applications in the real world
The technology that LLMs are based on had been around for decades until someone built one powerful enough to turn heads and change the world
The technology that LLMs are based on had been around for decades until someone built one powerful enough to turn heads and change the world
It would be cool to be able to “sponsor” other users to become members of a channel that you think they’d enjoy
Helps solve the problem of channel growth by tapping the people who are already in.
And it could be a source of serendipity for receiving users.
Helps solve the problem of channel growth by tapping the people who are already in.
And it could be a source of serendipity for receiving users.
Before the Industrial Revolution, you would work as an apprentice and learn how to make widgets.
The expectation was that you would eventually go off somewhere and begin making widgets on your own.
You learned the end-to-end process of widget making and selling
Most jobs these days provide little exposure to the end-to-end process, poorly equipping founders to go create their own.
I thought this was a big tech phenomenon but it’s actually a post industrialization consequence
The expectation was that you would eventually go off somewhere and begin making widgets on your own.
You learned the end-to-end process of widget making and selling
Most jobs these days provide little exposure to the end-to-end process, poorly equipping founders to go create their own.
I thought this was a big tech phenomenon but it’s actually a post industrialization consequence
Loading...
In chaos lies opportunity
Recently went down a water standards rabbit hole and would bet many PhDs would fail in answering what filters can effectively remove PFAS.
In order to filter out a health related contaminant, a filter needs to be NSF 53 *for that contaminant*.
Most manufacturers will just get an NSF 53 cert for one thing (like lead), then market their filters as if they've received the certs for all the baddies, like PFAS. Fooling a lot of people, even Consumer Reports (see pic)
What you want in general is a cert for each contaminant.
PFAS are special. Here the NSF level has varied over the years of testing, starting at 70 ppt, then moving to 20 ppt, which is still 5x the current 2024 standard, and >1,000x the EPA recommended standard of specific subclasses like PFOA.
Unfortunately that means the only solution right now is to understand from an engineering POV what types of filtration mechanisms are actually effective in removing PFAS.
Then buy a good filter of that type and cross your fingers.
In order to filter out a health related contaminant, a filter needs to be NSF 53 *for that contaminant*.
Most manufacturers will just get an NSF 53 cert for one thing (like lead), then market their filters as if they've received the certs for all the baddies, like PFAS. Fooling a lot of people, even Consumer Reports (see pic)
What you want in general is a cert for each contaminant.
PFAS are special. Here the NSF level has varied over the years of testing, starting at 70 ppt, then moving to 20 ppt, which is still 5x the current 2024 standard, and >1,000x the EPA recommended standard of specific subclasses like PFOA.
Unfortunately that means the only solution right now is to understand from an engineering POV what types of filtration mechanisms are actually effective in removing PFAS.
Then buy a good filter of that type and cross your fingers.
it's attestations all the way down
