People discuss their research in the open
Others respond with ideas and suggestions
That's why open research always wins at scale
TIGs flywheel means closed research can't compete
$TIG
A really interesting thing that has come out of my conversations with @Dr_JohnFletcher is the idea of leveraging the multimodal nature of the Gemini family to engage in visual reasoning for the differential geometry problem I’m working on
Trying something experiments now and will report the results back
most of the internet rewards whoever owns the biggest machine.
$TIG was built to reward something else.
some of its puzzles are set at the exact spot where they get as hard as they ever get, the place where adding more computers stops helping.
there, the only way up is to think of something smarter.
a small design choice with a big consequence. on this network, the people who win are the ones with better ideas.
We said we'd keep things moving. Here's another one.
tigstats, our live feed of network activity and events, just got an upgrade.
A fresh look and feel, new spotlights and features, plus more ways to follow what's happening across the entire ecosystem in real time.
Explore TIG's Challenges, Advances and latest Achievements. You can also find the progress being made across different areas of research, the breakthroughs pushing performance forward and the community that's contributing to it all.
Let us know what you think! https://t.co/Ahh7KJPMsy
@arthurmensch What happens when automated program search finds the "next Transformer", or optimiser, or attention mechanism, and kept private?
Every open weights model is obsolete... instantly
Models are not the problem. Its algorithms: https://t.co/BT4DSwJqT0
Without a new system of incentives, AI pushes science closed
The world’s best minds are recognising this
We saw it coming and designed a system to prevent it
the biggest routing puzzle on $TIG just got three times bigger.
until this week the delivery routing problem stopped at a thousand stops. now the network also asks for plans across two and three thousand, and two of the in-between sizes are gone. the code that earns here has to keep its quality as the map gets crowded.
a benchmark that never grows ends up rewarding code that is good at the benchmark. moving the bar up is how TIG keeps paying for work that gets closer to a real delivery day.
Open research and commercial value don't have to be at odds.
TIG's two-licence model gives researchers the freedom to share their work, while allowing businesses to use TIG algorithms commercially without giving away their proprietary data.
This reply from Gemini made me smile
If only it was just lawyers, economists, and politicians that completely misunderstood the most powerful technology ever created…
The first major use of the TIG Open Data License. Why does it matter, and what comes next?
Join @Dr_JohnFletcher and Phil on tomorrow's Community Call as they break down what this milestone means for TIG and why defining a new set of principles is essential to keeping AI and humanity free.
See you at 5PM GMT!
There has been a lot going on over the last few weeks ..
NGL its hard to not get stressed just by opening X
.. meanwhile $TIG is ignoring the noise and making good progress
The TIG thesis in action.
Eduardo Queiroga (UFBA) and Rafael Martinelli (PUC-Rio), both of whom help organise CVRPLib's vehicle-routing challenges, have built on Thibaut Vidal's HGS Advance algorithm, developed for our Vehicle Routing Challenge.
Their modified implementation is being used to generate best-known solutions for CVRPLib's upcoming Vehicle Routing Problem with Time Windows (VRPTW) Challenge. These are the solutions competitors will aim to beat.
Their work has also been published under the TIG Open Data License v3.0.
This is exactly what TIG was designed to make possible.
We want to create an environment where researchers can build on each other's contributions, rather than having every finding exist in isolation.
An algorithm developed for one challenge can become the basis for another, giving people a foundation to test new ideas, push its limits and make further discoveries.
This is how research can become cumulative. Less duplicated effort, more collaboration and faster progress.
We're not just rewarding individual achievements.
We're building a network where those breakthroughs can lead to many more, and then some.
More on the work -
Implementation: https://t.co/HlpDxyB3rV
Competition code to paper: https://t.co/TPhyrE3DG3
VRPTW Challenge overview: https://t.co/NB1axOXP5d