Sabitlenmiş Tweet
Maybe the real gains were the frens we made along the way $LINK
English
SatoshiSuperstar
8.7K posts
SatoshiSuperstar
@SatoshiSuperst1
What, Me Worry?
United States Katılım Eylül 2019
2.3K Takip Edilen722 Takipçiler
SatoshiSuperstar retweetledi
SatoshiSuperstar retweetledi
Just now finding out about @weatherchannel’s retro option, music and everything, and I am weeping openly. Feels so good to be back baby weather.com/retro/
English
SatoshiSuperstar retweetledi
“Our baby was born in Rivendell. That means she’s an elf and gets to vote now.”
English
English
SatoshiSuperstar retweetledi
SatoshiSuperstar retweetledi
Me to my wife when she asks me to take out the garbage and I do it immediately instead of sitting for another 5 minutes
English
SatoshiSuperstar retweetledi
I am paid a very high salary to design new ways to hurt innocent people. Please don't hurt me. I want to live. My job is to hurt your children. Please, don't do anything to me. I want to be safe. My job is to make your family sick. I want to live. Please, I deserve to live. Pleas
English
SatoshiSuperstar retweetledi
A Nintendo Switch running Linux running Cursor running Claude Code which wrote a script to launch Doom that was then reviewed by Codex which is running in Claude Code which is running in Cursor on a Nintendo Switch which is running Linux which is running Cursor which ran the script which launched Doom in Cursor.
English
SatoshiSuperstar retweetledi
We need more financial incentives for things
John Fletcher (𝔦, 𝔦)@Dr_JohnFletcher
Andrej, I’m John Fletcher. I have a PhD in mathematics and theoretical physics from Cambridge, and since 2016 I have been working full-time on the problem of how to coordinate untrusted distributed compute for algorithmic innovation. I listened to your No Priors conversation and recognised the architecture you were describing: commits that build on each other, computational asymmetry (hard to find, cheap to verify), an untrusted pool of workers collaborating through a blockchain-like structure. The result is The Innovation Game (TIG), which has been in continuous operation since mid-2024. The correspondence is so close that I thought it worth writing. The short version: roughly 7,000 Benchmarkers test algorithms submitted by Innovators by solving instances of asymmetric computational challenges (SAT, Vehicle Routing, Quadratic Knapsack, Vector Search, among others). This testing is "proof of work" in the technical sense of Dwork and Naor (1992). Innovators earn rewards proportional to adoption by the Benchmarkers. The repository of algorithms is open source (github.com/tig-foundation…). The system is already producing state-of-the-art results. For the Quadratic Knapsack Problem, 476 iterative submissions by independent contributors brought solution quality to a level that now exceeds methods published by Hochbaum et al. in the European Journal of Operational Research (2025). We are working with Thibaut Vidal (Polytechnique Montréal), who has submitted a state-of-the-art vehicle routing algorithm directly to TIG, and with Yuji Nakatsukasa (Oxford) and Dario Paccagnan (Imperial College London), among many others. One of TIG’s active challenges is directly relevant to your autoresearch work: an optimiser for neural network training (play.tig.foundation/challenges?cha…), where Innovators compete to develop an improved optimiser (see screenshot). One way in which TIG extends the vision is on the economic side. In our view, a monetary incentive is required, otherwise the open strand simply cannot compete at scale. TIG’s open source dual licensing model (designed by my co-founder Philip David, who was General Counsel at Arm Holdings for over a decade, and was the artchitect of ARMs licensing strategy) is intended to solve that problem. I expect we have each thought about parts of this that the other hasn’t. Happy to talk whenever suits. John Fletcher tig.foundation
English
SatoshiSuperstar retweetledi
English
SatoshiSuperstar retweetledi
Good evening my fellow Americans. I started this retarded war because I raped children on camera. My son in law is a Mossad agent and I met my wife through Jeffrey Epstein. My only honorable recourse is to ritually disembowel myself with a katana before you tonight
English
SatoshiSuperstar retweetledi
Hi Gagan, I don't think the ultimate triumph of open source AI is a safe assumption.
There is a degree of survivor bias here.
Was there a competitive open source alternative to Google web search?
I believe the same mechanism that gave Google a 25 year monopoly over search is present here too : x.com/Dr_JohnFletche…
Open source AI can win, but it's far from inevitable.
We need to make it happen. By providing the necessary incentives.
English
Best vid ever.
Im so excited for you all.
$TIG
ꋪꃅ0ꀸ꒒ (𝔦, 𝔦) 🏴☠️@Rh0DL
Watch this, and understand what is going to happen. $TIG
English
Extremely overlooked underrated game imo
Basically LOTR with FFX combat
PlayStation Nostalgia@PlayStalgiaX
The Lord of the Rings: The Third Age (2004)
English
SatoshiSuperstar retweetledi
Justin Drake’s striking observation points to a gap bigger than it first appears:
“From now on, assume state-of-the-art algorithms will be censored… A blackout in academic publications would be a tell-tale sign.”
@Dr_JohnFletcher builds from this toward a critical question:
What happens if algorithms go closed?
The answer affects not just cryptography, but the entirety of artificial intelligence.
The foundation of the current AI progress is not a dataset; it is an algorithm: the transformer architecture and the attention mechanism.
Publishing that code openly did not cost Google a competitive advantage; it earned them a historical legacy. But tools like AlphaEvolve are fundamentally changing the economics of algorithm discovery.
Open publication now places the discoverer at a material disadvantage.
This is precisely where the significance of John’s interpretation comes into view.
$TIG builds incentive mechanisms that make open publication the commercially optimal choice. The foundation of the system is a novel proof-of-work mechanism called OPoW (Optimisation Proof of Work): rather than performing cryptographically meaningless computations, miners here produce solutions to real-world optimisation problems, and solution quality directly determines reward. Roughly 7,000 Benchmarkers competitively test algorithms submitted by Innovators, continuously pressuring the performance frontier.
The scope of the challenges reflects the importance of the project.
Boolean Satisfiability, Vehicle Routing, and Quadratic Knapsack run alongside Vector Search, Hypergraph Partitioning, Neural Network Training Optimization, Job-Shop Scheduling, CUR Decomposition, Influence Maximisation, ZK Circuit Optimization, and Energy Grid Optimization, each either live or active on testnet.
Every challenge was designed in collaboration with leading domain experts: Prof. Thibaut Vidal (McGill University) for Vehicle Routing, Dr. Philipp Baumann (University of Bern) for Knapsack, Dr. Yuji Nakatsukasa (Oxford Mathematics) for CUR Decomposition, and Dr. Tasuku Soma (University of Tokyo) for Influence Maximisation and Hypergraph.
Measurable progress is already on record. On the Quadratic Knapsack challenge, 476 iterative contributions brought solution quality to a level that now exceeds results published by Hochbaum et al. in the European Journal of Operational Research (2025).
This is concrete evidence that an open, competitive structure can genuinely advance the peer-reviewed literature.
The licensing layer is the most strategically significant component.
Philip David, who served as General Counsel and IP Director at ARM Holdings for over a decade and was the architect of ARM’s intellectual property strategy through SoftBank’s acquisition, designed the open-source licensing model that bridges openness with commercial sustainability.
The model is built around a single premise:
That keeping an algorithm open should be more profitable than closing it at the moment of discovery.
At a moment when Shor’s algorithm is being pushed toward secrecy, keeping algorithms open requires that the economic logic of openness itself be redesigned. $TIG addresses this not as a technical problem, but at the level of mechanism design. That distinction changes everything.
John Fletcher (𝔦, 𝔦)@Dr_JohnFletcher
Justin, As you say.. “From now on, assume state-of-the-art algorithms will be censored. There may be self-censorship for moral or commercial reasons, or because of government pressure. A blackout in academic publications would be a tell-tale sign.” This, in my view, is the way that AI could go closed permanently. Not through hoarding data, or unavailability of hardware, but through SOTA algorithms going closed It’s easy to forget the current AI revolution was principally driven by the transformer architecture, which came from the attention mechanism: an algorithm. Algorithms are the highest leverage layer of the AI stack, and this leverage will only increase Algorithms have been under appreciated because, historically, they have been published openly. This is now changing, in large part due to AI-assisted algorithm discovery (see AlphaEvolve) which changes the economics of algorithm development so that open publication would but the discoverer at a significant competitive disadvantage The Innovation Game (TIG) was created to change these incentives, so that open publication is the commercially optimal route TIG has been in continuous operation since mid-2024. TIG has roughly 7,000 Benchmarkers test algorithms submitted by Innovators by solving instances of asymmetric computational challenges (SAT, Vehicle Routing, Quadratic Knapsack, Vector Search, among others) TIG is already producing state-of-the-art results. For the Quadratic Knapsack Problem, 476 iterative submissions brought solution quality to a level that now exceeds methods published by Hochbaum et al. in the European Journal of Operational Research (2025) Another challenge is an optimiser for neural network training (play.tig.foundation/challenges?cha…), where Innovators compete to develop an improved optimiser TIGs repository of algorithms is open source ( github.com/tig-foundation…). TIG works with some top the top experts in their fields including Thibaut Vidal (routing, explainable AI), Yuji Nakatsukasa (matries), Dario Paccagnan (game theory, mechanism design), among many others If this sounds familiar, it might be because Andrej @karpathy proposed a very similar vision in his recent No Priors interview with @saranormous See here for details x.com/Dr_JohnFletche… One way in which TIG extends Karpathy’s vision is on the economic side. In our view, a monetary incentive is required, otherwise the open strand simply cannot compete at scale TIG’s open source licensing model (designed by my co-founder Philip David, who was General Counsel at Arm Holdings for over a decade, and was the architect of ARMs licensing strategy) solves that problem Happy to discuss John Fletcher tig.foundation
English