Anders Sandberg

Academic jack-of-all-trades.

[Anders Sandberg @anderssandberg]

@LoganTCollins Yes! Science themed ties are so obviously done by people not familiar with the science. I want correct and artful patterns. There are so many good ones.

[Anders Sandberg @anderssandberg]

@DirkBruere AI development was a big surprise to everyone. I attended a seminar yesterday trying to make sense of post-2012 neural network overparameterization, and it is still a somewhat open research question.

[Anders Sandberg @anderssandberg]

I was just reading this 1925 book (because why not?) and randomly opened it to the page where Haldane dismisses nuclear weapons.

[Anders Sandberg @anderssandberg]

@jpsenescence Makes good sense.

[Anders Sandberg @anderssandberg]

@robinhanson Yes, it was great. It was wrong, but in productive ways. Understandibg where the method worked and why has been useful.

[Captain Pleasure, Andrés Gómez Emilsson @algekalipso]

Basically everyone on this site is overstating their case.

[Anders Sandberg @anderssandberg]

@gregeganSF My explanation is different departments in the bank. The UI people want to do a usable, or pretty, or fun (for them) interface. Then Compliance forces them to add the cookie popup. Marketing shows up and forces the ad - they got a high exec to sign off on it. Then Security...

[Anders Sandberg @anderssandberg]

@ben__h We need cyber-immune systems...

[Anders Sandberg @anderssandberg]

@QualiaNerd Yes, obviously "deep prosperity" involves superhappiness, or at least fixing most mental wants. When working on the postscarcity chapter in my book, it became clear that material PS < service PS < hedonic PS.

[Anders Sandberg @anderssandberg]

@Aella_Girl Hedging bets over vastly different discount rates sums to something like hyperbolic discounting.

I find it amusing that I have purely intellectual reasons to put more weight on shorter futures.

[Anders Sandberg @anderssandberg]

A piece of wisdom from professor Malcolm McMahon: "Anything giving you a factor ten is worth doing!"

[Anders Sandberg @anderssandberg]

Some good news in the Caribbean, which will likely turn postcolonial scholars into pretzels. https://t.co/j1hUj8lZRk

[Anders Sandberg @anderssandberg]

This is promising. Plus yet another reason to use Signal.

[Pushmeet Kohli @pushmeet]

Announcing a major milestone achieved by my team @GoogleDeepMind: we have proven key security properties of the machine code underlying the Signal messaging protocol. This is the first time the actual production code of a widely deployed protocol has been formally verified – by any method, with or without AI – and at this depth.

More than a research breakthrough, it’s a prime example of how frontier AI can be a force multiplier to counter the speed and scale of modern cyber threats.

Formal verification offers a higher standard of security; however, it’s extremely complex and time consuming, which is why in many cases it’s considered to be impractical. Similar efforts could take teams decades of combined effort. We did it in a month’s time.

Using AlphaProof Code, Google DeepMind's new agentic framework for formal verification that runs on LLMs including Gemini Carbon, we verified the Signal protocol down to the actual code running on devices.  This milestone lays the groundwork for a future in which vital systems, from medical device firmware to power grid software, ship with mathematical proof that their protections work.

Reaching that future will take continued research from our community. In that spirit, we're open-sourcing our formal specifications and verification artifacts so others can build on this work.

On a personal note, this moment is of deep significance to me. I started my own research journey more than two decades ago as an intern working on checking the consistency of complex software systems. Seeing this grand challenge made practical with modern AI is extraordinary.

I'm extremely proud of my team members: Swarat Chaudhuri, Andrew Ferraiuolo, Anton Kovsharov, Tej Chajed, Sam Schlesinger,  Karn Seth, Jonathan Katz, and Charles Sutton for taking on this grand challenge and delivering, and grateful to the Signal team for their partnership.

https://t.co/kN1eRYQ8fF

[Anders Sandberg @anderssandberg]

@S_OhEigeartaigh One can also be the embarrassing aunt at the family dinner: "Look at you! Wow! I remember when you were a tiny undergraduate!"

[Anders Sandberg @anderssandberg]

@S_OhEigeartaigh Same here. One of the best things with academic hubs is that you get to meet people who go off and change the world when they were young.

[E11 Bio @E11BIO]

Love hydrogels, microscopes & hard problems? 🧠

We are hiring an Expansion Microscopy Scientist to scale sample production towards whole-brain connectomics.

📍 Alameda, CA
🔗 Apply here: https://t.co/TLYCtdCK80

#Hiring #Neuroscience #SfN26 #SfN @SfNtweets

[Anders Sandberg @anderssandberg]

@marcodecunha1 @dunoqft Yes, we are using very wide priors - but as we argue in the text and appendices, these are actually what is supported by what we know, and somebody arguing for a mere OoM or two uncertainty for the later parameters is being much bolder than the evidence allows.

[Anders Sandberg @anderssandberg]

@robinhanson Generally, do we have any examples of boards doing testing of their executives for fit after they are appointed?

[Logan Thrasher Collins @LoganTCollins]

A brief plea for the public to embrace a more clearheaded bioethical framework around longevity Substack link: https://t.co/ulAWPYsBIF

[Anders Sandberg @anderssandberg]

@LilyEricksonMN @thesupermannx The question is what the overall aim is. It can make sense to be quiet and cold. Or build a 1000 AU Dyson around a supermassive black hole accretion disk with neutron star flywheels and whatnot. Stellar Dysons are plausible stepping stones, but I agree unlikely the final word.

[Anders Sandberg @anderssandberg]

@LilyEricksonMN @thesupermannx Gravitational lensing only works for dense objects; these would be more occultation or re-radiated energy forming technosignature spectra. If you see signs of weird optical coatings, tungsten, tantalum hafnium carbide or beryllium, something is going on.