AI Sandbagging – Computerphile

From Computerphile. Following the theme of AI research and safety, Aric Floyd talks about how some Large Language Models might follow the all too human trait of sandbagging – "lying" about their true capabilities. AI Sandbagging Paper: https://www.apolloresearch.ai/research/scheming-reasoning-evaluations Computerphile is supported by Jane Street. Learn more about them (and exciting career opportunities) at: https://jane-st.co/computerphile This…

Subroutines in Low Level Code – Computerphile

From Computerphile. Bashing out low-level code, it can be annoying to re-type the same commands over and over when you need to repeat a routine. Matt Godbolt explains how we can save frequently used code as a subroutine and reuse whenever its needed. Matt Godbolt is known as the creator of Compiler Explorer, among other…

Is this AI’s Version of Moore’s Law? – Computerphile

From Computerphile. Research suggests the rate at which AI is able to stay ‘on task’ is doubling every seven months. Is this the AI equivalent of Moore’s Law? – Sydney Von Arx works on adversarial stress-testing at METR (Model Evaluation & Threat Research.) Computerphile is supported by Jane Street. Learn more about them (and exciting…

Shortest Path Algorithm Problem – Computerphile

From Computerphile. A seemingly simple problem that’s "in general" incredibly difficult! CEO of Redwood Research Buck Shlegeris explains his favourite algorithmic fact! Buck wants to thank his friend Peter Schmidt-Nielsen for telling him this great fact. More on the sums of square roots problem: https://cstheory.stackexchange.com/questions/79/problems-between-p-and-npc#4010 Computerphile is supported by Jane Street. Learn more about them…

Ai Will Try to Cheat & Escape (aka Rob Miles was Right!) – Computerphile

From Computerphile. As Large Language Models improve, the tokens they predict form ever more complicated and nuanced outcomes. Rob Miles and Ryan Greenblatt discuss "Alignment Faking" a paper Ryan’s team created – ideas about which Rob made a series of videos on Computerphile in 2017. The Alignment Faking paper: https://tinyurl.com/C-Paper-AlignmentFaking Ryan Greenblatt is chief scientist…

Carbon: (C++)++ ? Can Carbon De-Throne C++? – Computerphile

From Computerphile. This video features Dr Valerio Guiffrida. Full details plus info on the NVIDIA GTC offer can be found below in the full description ↓↓↓ Carbon is a brand new programming language still under development. It aims to dethrone C++ by being interoperable & compatible in both directions. Dr Guiffreida from the University of…

No Regrets – What Happens to AI Beyond Generative? – Computerphile

From Computerphile. Discussing ideas of what happens after Generative AI plateaus, Dr Jakob Foerster is based at the University of Oxford. Try the FLAIR simulator here: https://kinetix-env.github.io/gallery.html?editor=true Papers: https://arxiv.org/pdf/2410.23208 https://arxiv.org/abs/2308.13289 Computerphile is supported by Jane Street. Learn more about them (and exciting career opportunities) at: https://jane-st.co/computerphile This video was filmed and edited by Sean Riley.…

Finding The Slope Algorithm (Forward Mode Automatic Differentiation) – Computerphile

From Computerphile. The algorithm for differentiation relies on some pretty obscure mathematics, but it works! Mark Williams demonstrates Forward Mode Automatic Differentiation. Computerphile is supported by Jane Street. Learn more about them (and exciting career opportunities) at: https://jane-st.co/computerphile This video was filmed and edited by Sean Riley. Computerphile is a sister project to Brady Haran’s…

Software Engineering for Quantum Computing – Computerphile

From Computerphile. As quantum devices become more accessible, there’s a whole area of software engineering opening up. Mohammad Mousavi specialises in software engineering for quantum computing. Mohammad Reza Mousavi is a professor of Software Engineering at King’s College London & one of the centre directors at King’s Quantum. Computerphile is supported by Jane Street. Learn…

Using Bayesian Approaches & Sausage Plots to Improve Machine Learning – Computerphile

From Computerphile. Bayesian logic is already helping to improve Machine Learning results using statistical models. Professor Mike Osborne drew us some sausage plots to explain the idea. Mike Osborne is Professor of Machine Learning at the University of Oxford – Find out more here: http://www.robots.ox.ac.uk/~mosb? Computerphile is supported by Jane Street. Learn more about them…

Malleable Encryption – Computerphile

From Computerphile. Malleable encryption means you can flip a bit in the encrypted message and the corresponding bit is flipped in the unencrypted plain text. Dr Tim Muller gives us some examples. nb This video is a reupload due to an audio problem with the first version. For those interested there were tones at approx…