Target 3: Taking It To The Next Level

 |   |  MIRI Strategy

One week ago, we hit our first fundraising target. I’m thrilled to announce that we’re now closing in on our second target: our fundraising total passed $400,000 today!

As we approach target number two, we’re already taking active steps to grow our team. Jessica Taylor joined our core research team on August 1; another research fellow will be coming on in September; and a third researcher has just signed on to join our team in the near future — details forthcoming. These three new recruits will increase the size of our team to six full-time researchers.

We’re courting a few other researchers who may be able to join us later in the year. Meanwhile, we’re running a workshop on logical uncertainty, and we’ve started onboarding a new intern with the aim of helping us with our writing bottleneck.

We’re already growing quickly — but we could still make use of additional funds to pursue a much more ambitious growth plan. Given that we’re only halfway through our fundraiser, this is a good time to start thinking big.

At present, we’re recruiting primarily from a small but dedicated pool of mathematicians and computer scientists who come to us on their own initiative. If our fundraiser successfully passes target number two, any further funds will enable us to pivot toward recruiting top talent more broadly — including highly qualified mathematicians and computer scientists who have never heard of us before.

We have a strong pitch: we’re working on some of the most interesting and important problems in the world, on a research topic which is still in its infancy. There is lots of low-hanging fruit to be picked, and the first papers on these topics will end up defining this new paradigm of research. Researchers at MIRI have a rare opportunity to make groundbreaking discoveries that may play a critical role in AI progress over the next few decades.

Moreover, MIRI researchers don’t have to teach classes, and they aren’t under a “publish or perish” imperative. Their job is just to focus on the most important technical problems they can identify, while leaving the mundane inconveniences of academic research to our operations team. When we make it our priority to recruit the world’s top math talent, we’ll be able to put together a pretty tempting offer!

This is what we’ll do at funding target number three: Take MIRI’s growth to the next level. At this level, we’ll start stepping up our recruitment efforts to build our AI alignment dream team.

Read more »

When AI Accelerates AI

 |   |  Analysis

Last week, Nate Soares outlined his case for prioritizing long-term AI safety work:

1. Humans have a fairly general ability to make scientific and technological progress. The evolved cognitive faculties that make us good at organic chemistry overlap heavily with the evolved cognitive faculties that make us good at economics, which overlap heavily with the faculties that make us good at software engineering, etc.

2. AI systems will eventually strongly outperform humans in the relevant science/technology skills. To the extent these faculties are also directly or indirectly useful for social reasoning, long-term planning, introspection, etc., sufficiently powerful and general scientific reasoners should be able to strongly outperform humans in arbitrary cognitive tasks.

3. AI systems that are much better than humans at science, technology, and related cognitive abilities would have much more power and influence than humans. If such systems are created, their decisions and goals will have a decisive impact on the future.

4. By default, smarter-than-human AI technology will be harmful rather than beneficial. Specifically, it will be harmful if we exclusively work on improving the scientific capability of AI agents and neglect technical work that is specifically focused on safety requirements.

To which I would add:

  • Intelligent, autonomous, and adaptive systems are already challenging to verify and validate; smarter-than-human scientific reasoners present us with extreme versions of the same challenges.
  • Smarter-than-human systems would also introduce qualitatively new risks that can’t be readily understood in terms of our models of human agents or narrowly intelligent programs.

None of this, however, tells us when smarter-than-human AI will be developed. Soares has argued that we are likely to be able to make early progress on AI safety questions; but the earlier we start, the larger is the risk that we misdirect our efforts. Why not wait until human-equivalent decision-making machines are closer at hand before focusing our efforts on safety research?

One reason to start early is that the costs of starting too late are much worse than the costs of starting too early. Early work can also help attract more researchers to this area, and give us better models of alternative approaches. Here, however, I want to focus on a different reason to start work early: the concern that a number of factors may accelerate the development of smarter-than-human AI.

AI speedup thesis. AI systems that can match humans in scientific and technological ability will probably be the cause and/or effect of a period of unusually rapid improvement in AI capabilities.

Read more »

August 2015 Newsletter

 |   |  Newsletters

Research updates

General updates

News and links

  • We’re at day three of the Effective Altruism Global conference! You can watch a selection of talks on the livestream.
  • Thousands sign an open letter by the Future of Life Institute advocating “a ban on offensive autonomous weapons beyond meaningful human control.”

A new MIRI FAQ, and other announcements

 |   |  News

MIRI is at Effective Altruism Global! A number of the talks can be watched online at the EA Global Livestream.

We have a new MIRI Frequently Asked Questions page, which we’ll be expanding as we continue getting new questions over the next four weeks. Questions covered so far include “Why is safety important for smarter-than-human AI?” and “Do researchers think AI is imminent?

We’ve also been updating other pages on our website. About MIRI now functions as a short introduction to our mission, and Get Involved has a new consolidated application form for people who want to collaborate with us on our research program.

Finally, an announcement: just two weeks into our six-week fundraiser, we have hit our first major fundraising target! We extend our thanks to the donors who got us here so quickly. Thanks to you, we now have the funds to expand our core research team to 6–8 people for the coming year.

New donations we receive at will now go toward our second target: “Accelerated Growth.” If we hit this second target ($500k total), we will be able to expand to a ten-person core team and take on a number of important new projects. More details on our plans if we hit our first two fundraiser targets: Growing MIRI.

MIRI’s Approach

 |   |  Analysis

MIRI’s mission is “to ensure that the creation of smarter-than-human artificial intelligence has a positive impact.” How can we ensure any such thing? It’s a daunting task, especially given that we don’t have any smarter-than-human machines to work with at the moment. In the previous post I discussed four background claims that motivate our mission; in this post I will describe our approach to addressing the challenge.

This challenge is sizeable, and we can only tackle a portion of the problem. For this reason, we specialize. Our two biggest specializing assumptions are as follows:

We focus on scenarios where smarter-than-human machine intelligence is first created in de novo software systems (as opposed to, say, brain emulations).

This is in part because it seems difficult to get all the way to brain emulation before someone reverse-engineers the algorithms used by the brain and uses them in a software system, and in part because we expect that any highly reliable AI system will need to have at least some components built from the ground up for safety and transparency. Nevertheless, it is quite plausible that early superintelligent systems will not be human-designed software, and I strongly endorse research programs that focus on reducing risks along the other pathways.

We specialize almost entirely in technical research.

We select our researchers for their proficiency in mathematics and computer science, rather than forecasting expertise or political acumen. I stress that this is only one part of the puzzle: figuring out how to build the right system is useless if the right system does not in fact get built, and ensuring AI has a positive impact is not simply a technical problem. It is also a global coordination problem, in the face of short-term incentives to cut corners. Addressing these non-technical challenges is an important task that we do not focus on.

In short, MIRI does technical research to ensure that de novo AI software systems will have a positive impact. We do not further discriminate between different types of AI software systems, nor do we make strong claims about exactly how quickly we expect AI systems to attain superintelligence. Rather, our current approach is to select open problems using the following question:

What would we still be unable to solve, even if the challenge were far simpler?

For example, we might study AI alignment problems that we could not solve even if we had lots of computing power and very simple goals.

We then filter on problems that are (1) tractable, in the sense that we can do productive mathematical research on them today; (2) uncrowded, in the sense that the problems are not likely to be addressed during normal capabilities research; and (3) critical, in the sense that they could not be safely delegated to a machine unless we had first solved them ourselves. (Since the goal is to design intelligent machines, there are many technical problems that we can expect to eventually delegate to those machines. But it is difficult to trust an unreliable reasoner with the task of designing reliable reasoning!)

These three filters are usually uncontroversial. The controversial claim here is that the above question — “what would we be unable to solve, even if the challenge were simpler?” — is a generator of open technical problems for which solutions will help us design safer and more reliable AI software in the future, regardless of their architecture. The rest of this post is dedicated to justifying this claim, and describing the reasoning behind it.

Read more »

Four Background Claims

 |   |  Analysis

MIRI’s mission is to ensure that the creation of smarter-than-human artificial intelligence has a positive impact. Why is this mission important, and why do we think that there’s work we can do today to help ensure any such thing?

In this post and my next one, I’ll try to answer those questions. This post will lay out what I see as the four most important premises underlying our mission. Related posts include Eliezer Yudkowsky’s “Five Theses” and Luke Muehlhauser’s “Why MIRI?”; this is my attempt to make explicit the claims that are in the background whenever I assert that our mission is of critical importance.


Claim #1: Humans have a very general ability to solve problems and achieve goals across diverse domains.

We call this ability “intelligence,” or “general intelligence.” This isn’t a formal definition — if we knew exactly what general intelligence was, we’d be better able to program it into a computer — but we do think that there’s a real phenomenon of general intelligence that we cannot yet replicate in code.

Alternative view: There is no such thing as general intelligence. Instead, humans have a collection of disparate special-purpose modules. Computers will keep getting better at narrowly defined tasks such as chess or driving, but at no point will they acquire “generality” and become significantly more useful, because there is no generality to acquire. (Robin Hanson has argued for versions of this position.)

Short response: I find the “disparate modules” hypothesis implausible in light of how readily humans can gain mastery in domains that are utterly foreign to our ancestors. That’s not to say that general intelligence is some irreducible occult property; it presumably comprises a number of different cognitive faculties and the interactions between them. The whole, however, has the effect of making humans much more cognitively versatile and adaptable than (say) chimpanzees.

Why this claim matters: Humans have achieved a dominant position over other species not by being stronger or more agile, but by being more intelligent. If some key part of this general intelligence was able to evolve in the few million years since our common ancestor with chimpanzees lived, this suggests there may exist a relatively short list of key insights that would allow human engineers to build powerful generally intelligent AI systems.

Further reading: Salamon et al., “How Intelligible is Intelligence?
Read more »

Why Now Matters

 |   |  MIRI Strategy

I’m often asked whether donations now are more important than donations later. Allow me to deliver an emphatic yes: I currently expect that donations to MIRI today are worth much more than donations to MIRI in five years. As things stand, I would very likely take $10M today over $20M in five years.

That’s a bold statement, and there are a few different reasons for this. First and foremost, there is a decent chance that some very big funders will start entering the AI alignment field over the course of the next five years. It looks like the NSF may start to fund AI safety research, and Stuart Russell has already received some money from DARPA to work on value alignment. It’s quite possible that in a few years’ time significant public funding will be flowing into this field.

(It’s also quite possible that it won’t, or that the funding will go to all the wrong places, as was the case with funding for nanotechnology. But if I had to bet, I would bet that it’s going to be much easier to find funding for AI alignment research in five years’ time).

In other words, the funding bottleneck is loosening — but it isn’t loose yet.

We don’t presently have the funding to grow as fast as we could over the coming months, or to run all the important research programs we have planned. At our current funding level, the research team can grow at a steady pace — but we could get much more done over the course of the next few years if we had the money to grow as fast as is healthy.

Which brings me to the second reason why funding now is probably much more important than funding later: because growth now is much more valuable than growth later.

There’s an idea picking up traction in the field of AI: instead of focusing only on increasing the capabilities of intelligent systems, it is important to also ensure that we know how to build beneficial intelligent systems. Support is growing for a new paradigm within AI that seriously considers the long-term effects of research programs, rather than just the immediate effects. Years down the line, these ideas may seem obvious, and the AI community’s response to these challenges may be in full swing. Right now, however, there is relatively little consensus on how to approach these issues — which leaves room for researchers today to help determine the field’s future direction.

People at MIRI have been thinking about these problems for a long time, and that puts us in an unusually good position to influence the field of AI and ensure that some of the growing concern is directed towards long-term issues in addition to shorter-term ones. We can, for example, help avert a scenario where all the attention and interest generated by Musk, Bostrom, and others gets channeled into short-term projects (e.g., making drones and driverless cars safer) without any consideration for long-term risks that are less well-understood.

It’s likely that MIRI will scale up substantially at some point; but if that process begins in 2018 rather than 2015, it is plausible that we will have already missed out on a number of big opportunities.

The alignment research program within AI is just now getting started in earnest, and it may even be funding-saturated in a few years’ time. But it’s nowhere near funding-saturated today, and waiting five or ten years to begin seriously ramping up our growth would likely give us far fewer opportunities to shape the methodology and research agenda within this new AI paradigm. The projects MIRI takes on today can make a big difference years down the line, and supporting us today will drastically affect how much we can do quickly. Now matters.


Targets 1 and 2: Growing MIRI

 |   |  MIRI Strategy

Momentum is picking up in the domain of AI safety engineering. MIRI needs to grow fast if it’s going to remain at the forefront of this new paradigm in AI research. To that end, we’re kicking off our 2015 Summer Fundraiser!

Rather than naming a single funding target, we’ve decided to lay out the activities we could pursue at different funding levels and let you, our donors, decide how quickly we can grow. In this post, I’ll describe what happens if we hit our first two fundraising targets: $250,000 (“continued growth”) and $500,000 (“accelerated growth”).

Read more »