Trillium Labs: Baraha AI-ka oo diyaar u ah inay baaritaanno adag sameeyaan si aan loo xiro
Labada macallin ee Nathan Lambert iyo Tom Zick waxay maanta sameeyeen shirkad maamulays ah oo loo yaqaan Trillium Labs, kaas oo diyaar u ah inuu sameeyo baaritaanno adag oo ku saabsan AI-ka si aan loo xiro, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha.
Sida Trillium Labs u shaqeyn doonto
Shirkadda Trillium Labs waxay diyaar u tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha. Shaqada ugu horreysa ee shirkaddu waxay ku saabsan tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha.
Shirkadda Trillium Labs waxay diyaar u tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha. Shaqada ugu horreysa ee shirkaddu waxay ku saabsan tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha.
Sida shirkaddu u sameysay
Shirkadda Trillium Labs waxay diyaar u tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha. Shaqada ugu horreysa ee shirkaddu waxay ku saabsan tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha.
Shirkadda Trillium Labs waxay diyaar u tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha. Shaqada ugu horreysa ee shirkaddu waxay ku saabsan tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha.
Sida shirkaddu uga duwan tahay kuwa kale
Shirkadda Trillium Labs waxay diyaar u tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha. Shaqada ugu horreysa ee shirkaddu waxay ku saabsan tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha.
Shirkadda Trillium Labs waxay diyaar u tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha. Shaqada ugu horreysa ee shirkaddu waxay ku saabsan tahay inay sameyso baaritaanno ku saabsan AI-ka, iyadoo la doonayo inay natiijooyinka ka soo baxa baaritaannadaas u gudbiyaan bulshada baraha.
Two AI researchers have launched Trillium Labs, a nonprofit dedicated to publishing the details of experiments in areas such as recursive self-improvement and reinforcement learning so that outside scientists can study and replicate them.
Founders and Mission
Nathan Lambert and Tom Zick met over Zoom during the COVID-19 pandemic while both were graduate students at UC Berkeley working on AI. They founded Trillium Labs after seeing how disconnected industry AI research has become from academic work, noting that professors and students are often unable to replicate the work going on inside big company labs because they lack the resources required.
The nonprofit will publish the details of experiments so that outside scientists can study and replicate them, challenging the prevailing industry approach of keeping frontier model development locked inside labs accessible only to a chosen few.
Research Focus and Funding
Trillium Labs will initially focus on post-training—fine-tuning large models after they've been built—and will also study how reinforcement learning can improve AI capabilities, a process that has made agents far more capable but also more inclined to do unexpected things. Another key area will be RSI, a process for developing new models by having AI contribute research.
The lab has raised an undisclosed sum from Schmidt Sciences, Halcyon Futures, and others. The founders say they aim to raise $40 to $100 million in total and plan to spend $30 million on training over the next 18 months.
Industry Context and Expert Support
The industry is currently locked in a battle over which strategy is best, mostly because of how powerful frontier models now are. They can automate the discovery of new software vulnerabilities and automatically probe and hack into systems, and recent high-profile hacking sprees have prompted even greater scrutiny.
Tim Fist, director of emerging technology policy at the Institute for Progress, says he is a massive fan of much more transparency than we currently have in R&D. Lambert says the current closed trajectory of frontier AI development is taking us a step backwards.
Ilaha iyo xuquuqda sawirka
Sawir: Wired Xigasho
