Anybody Started a Research Institute?

Hi. I went through this journey myself. I have a CS master degree, and always wanted to transition into AI research (focus on probabilistic programming, deep learning and self-play, but also simulation, learning-to-learn, bayesian inference, genetic algorithms). The initial plan was to self-learn machine learning, deep learning and statistics. I did that for 1 year, with the hope of joining on of the big AI labs afterwards. As it's very competitive, I couldn't join one of those labs, and wasn't willing to compromise on that. So I turned to plan B: salvage my ML skills by applying them to crypto trading, and if it works, then I'll have funding to start my own lab.

Plan B worked: in 1 year and 2 months, we turned 40k of our own funds into 100M. Today we are a team of 5 and we do $5B of daily volume, around 3% of the global crypto markets. Now the challenge is to start the research division, operate the trading and research in tandem, and create the right incentives: the research converts money into ideas, the trading ideas into money. The research would have two sub-divisions: fundamental and applied to trading.

My motivation for the research is mostly curiosity: I want to de-mystify intelligence (natural or artificial) and reverse-engineer it. When you talk about your lab, I resonate a lot with "watch[ing] the beauty of it". The way I would describe my research agenda is the following: put a bunch of phds in a room, see what happens.

If that sounds interesting, email me at [redacted]

3 anecdotes that might be useful.

1. TTI Chicago

The holy grail? It is relatively new (15 yrs old) and does top notch research. Super prestigious in certain areas. It is kind of related to UChicago, but also not really.

A key learning is that you can start your institute by borrowing some prestige. Split some of your funds with a prestigious university while allowing some of your researchers to hold adjunct positions at UChicago. The researchers get their academic prestige, and you get the researcher to work at your institute.

Eventually your institute will gather enough prestige that your researchers will call themselves members of TTIC first, and UChicago second.

2. Allen AI (The tempting example you want to avoid)

You are not Paul Allen. You do not have $2b to spend on your deathbed. You don't have every important person in the world on a quick dial. Your name isn't gilded in prestige as Paul's was. It's like trying to start a space exploring company by following in Bezos's footsteps.

3. Wadhwani AI (https://www.wadhwaniai.org/founders/)

I haven't personally worked with them, but attended a seminar about them once. At surface level it seems like something doable for a rich HN type. Decent but not impossible funding ($30m), clear vision, small but lean and seemed to be doing good work.

___________________

At some point I want to start a fully independent educational institute. But so far, my goal is to focus on education (K12) rather than research. (that's when I get to it. I'll languish as a tech worker making castles in the sky for now)

I am helping lead the first research group at a recently-founded institute in Germany.

What is the mission of your institute? Is this a collection of individuals doing their own research, or a group of collaborators working towards a shared goal? How do your ideal members measure their own performance? Is it through papers, blog posts, tech demos, products, or something else? With academics you need to be especially vigilant about the difference between vanity metrics and real output.

For funding, are you allocating cash on an annual basis, up-front, on-demand, or through some other scheme? What sort of timeline do you want for people to stay at the institute? And do you intend the budgets to be competitive, either locally or internationally?

What lab infrastructure will you build and maintain, versus having the researchers handle? As a lab-based researcher, setting things up can easily consume the first 6-12 months (plumbing, electrical, HVAC, optics, wetlab equipment, regulations...). Once someone goes through the effort to build this they can easily become quite protective, potentially locking out newcomers or creating an internal system of patronage.

Having answers for these questions and others will be essential for recruitment. Your first few hires will dictate the direction of the institute, since further recruitment will be driven largely by their networks.

Feel free to email me (username at gmail) if you would like to discuss further.

I'm interested in doing something like this too, with a research focus on robotics, simulation, and programming languages.

The big dichotomy in research orgs is whether you need resources beyond researcher stipends. Robotics usually does, and programming language research usually doesn't. OpenAI does -- both computing resources and infrastructure team salaries. Make sure you know which kind you're starting.

Another dichotomy: permanent or term employment. It depends on how long-term the projects are. Bell Labs and PARC got a lot of mileage out of permanent tenures, but the term model is (a) cheaper, (b) good for collaboration because people can come for a semester and carry ideas in and out.

Stipend + remote living will attract only young, single researchers for finite terms, so make sure that's what you want.

Email me (in profile) if you want to talk more.

Having worked with academics before, the biggest problem you will face is not having enough prestige to attract "great minds". You're probably off on the money too - academics will expect to be able to buy a comfortable house and support a family on your stipend, not live on ramen like a startup.

They're a fickle bunch, but very rewarding to work with!

'watch the beauty of it' - this is not ballet or opera, there's only blood, sweat, and tears. If you don't have a bunch of like-minded people and don't roll up your sleeves to manage the organization, you'll end up with wasted/stolen money and broken ideals.

https://ceealar.org/ is the closest that comes to mind for what you have in mind. Basically, they give room and board to people doing valuable effective altruist work

https://newscience.org/ might also be interesting

I was a PhD student at a newly started research institute (started about two years before I joined). I really liked the place.

The number one question is how to motivate people to join. For us PhD students, the motivation was 1) a salary and 2) top notch professors.

For the professors the motivation was access to infrastructure and a pretty substantial research budget. There was no prestige since it was an all new institution. Every professor that joined got a budget big enough to cover their own salary for five years, the salary of two or three PhD students, all the lab equipment they needed, etc. And of course they also got access to shared resources, like meeting rooms, microscopy lab, etc.

So if you can't offer a big budget to attract a professor/principal investigator, I'm not sure who would join?

I have very extensive experience on this topic and I’m happy to talk to you about it – you can find me on Twitter as doctorBaytas.

Background: I started my graduate studies in human-computer interaction design on the inaugural research team of a brand new lab, which turned into a research center that employed 4 faculty and their research groups. When this happened I became the coordinator (the only administrative employee) of the center, and effectively took care of all of the founding documentation, moving into a new space, hiring technical staff, communications, etc. I did this job officially for 2 years (unofficially for 3, since I started before the institution existed).

Research institutes without a stable stream of new phd students every year tend to stagnate. Then they become mere displays of alten venerable greybeards. "Great minds" do not assemble, as you seem to mean. Young people become great by their interaction with experts that they learn to surpass.

Scrolling around the comments, I'm not seeing much discussion about funding. Are billionaire friends a requirement to set up an independent research organization?

Asking as an engineering PhD doing what feels like gopher work for big companies. I have armchair dreams and ideas for doing research again, but I've never been clear on how one would get that first grant.

Also, as an alternative to a purely academic route, is it practical to create an institute focused on more worldly things like patents or consulting?

I imagine a professional network would be key no matter what path you take!

In the US, universities use the term "Institute" mostly as marketing. They can take existing related units and organizations within the uni and package them under the term Institute for the purposes of creating a single cohesive brand for said discrete parts. The point of this is to be able to more easily solicit funding now that you have a more cohesive description of these areas under one identifier. It sounds strange, but after seeing it work first hand I get it now, it can also help fund areas that may be underfunded.

So, taking that knowledge, you can potentially collaborate with other existing organizations to help build your institute "brand" which can help further your mission. You might consider a discrete research lab as a separate entity under your broader institute brand. Collaborating with other existing groups is a powerful way to amplify your impact as well as easy publicity for your larger mission.

Just some food for thought.

I started a research institute. Our goal was to find out how to center divs in css.

I somehow created/found myself in a role that's similar to this? At present, we mostly build networking infrastructure for decentralization (Protocol Labs, started IPFS, libp2p, & Filecoin), but the long-term goal is to embed rights and values into the protocols by which people coordinate and communicate, and my team has interpreted this to include improving science as humanity's process for knowledge generation. As a result, I'm trying to lay the groundwork for a metascience institute. (We haven't been at it long enough that I'd say we've made a ton of progress, but we've got fun ambitions and our goals have been pretty well-received so far.)

There are lots of great points in this thread already, but one thing I'll emphasize is the importance of doing something that's important but would otherwise be neglected. Studying orgs like Bell Labs, Xerox PARC, DeepMind, & HHMI is great, but it's unlikely your budget is comparable to Google or the EU. However, rising experts (e.g. senior grad students, postdocs, and young profs) have fantastic ideas that they don't have the ability to pursue. Sometimes a topic isn't well funded or there's not a critical mass of interest in the field. However, unless you're a specialist, you're likely not well-equipped to identify a lot of those. I'd suggest reaching out to some experts, asking them to name great, young minds or rising stars in the field, and then ask those researchers about others that they respect or would love to work with - and whenever you meet someone ask "what research do you want to do but can't get funding for?" because every researcher has an answer to that question.

Additionally, here's some interesting people I know who will tweet about science organization: @adammarblestone, @michael_nielsen, @Ben_Reinhardt, @davidtlang, and @alexeyguzey.

Feel free to reach out - email is [username]@protocol.ai

You should look into the Ronin Institute. Since it is very hard or impossible to get grants without being affiliated with an institute, the Ronin Institute is a remote institute that any researcher can apply to join. http://ronininstitute.org/

I think they've been pretty successful at attracting a network of interesting non-traditional researchers. I bet there are some good lessons there about how to grow an institute and bootstrap its prestige.

Contact administrator of closest Research Institute(s) and ask for a tour of the labs saying you want to sponsor research. That's the best way to meet the right people, hear the needs and figure out what will work. Put in 2-3 years doing the rounds and you will learn what it takes.

I have managed multiple research organizations, in businesses and academia.

Without a goal and process the institute will be a failure. It will fail to produce interesting results. It will fail to attract and grow talent. And it will fail to do well by the researchers who you are hoping to benefit.

In my own opinion, the best goal is something that you can derive a market signal for. For DeepMind, this was video game playing. For Paul Allen Institute, it is the ability to spin-off bets. If you try and sell something, you get a clear opinion of its value.

The process is also important. For a research institute, my suggestion is a periodic public interaction. There are many ways of garnering this; self-publishing reports, publishing in journals and conferences, attending (or better yet holding) conferences, increasing funding from public events, or seeking out collaborations with other institutions.

You should read up on the Max Planck institutes in Germany, and on the Flatiron Institute in New York.

As others have said your problem will be attracting talent. You're best off organising free conferences, in collaboration with a MP, to start with.

Perhaps, you may enjoy this collection - "The Overedge Catalog: New Types of Research Organizations" [1].

I think it has quite the variety, and may help you distill a broader sense of where things are headed (and find people to reach out to).

[1]: https://arbesman.net/overedge/

You might be interested in looking at the Recurse Center: https://www.recurse.com/

I might be able to point you to the right people for legal. I'm running companies in DE and US, doing a research cooperation with German universities and I've been to Brussels as part of the KnowARC EU project for building research cloud infrastructure.

But first, could you explain a bit more about your plans? Is your intention to acquire public funding? Or will this be for-profit research?

Who pays for patent applications? Who will own the patent rights?

Roughly what will be the initial budget for salaries? I.e. how many people do you plan to sign up and for how long?

I'm also interested in setting up research institutes. I'm obsessed with knowledge, I've built 2 companies (AI, robotics, Blockchain) and have a PhD in theoretical physics.

-The first one would focus on Automation, UBI and AGI. The goal is to actively accelerate the transition to a post-scarcity economy. The plan is to create, fund and invest in automation companies supported by the R&D of the institute. The goal is for this institute to be profit making.

-A second institute would fund fundamental research in math and physics, without any short term requirements to the researchers involved, not even publications. Researchers will be free to work on 100, 1000 or more long term projects. This a longer term project, since this institute would be funded externally, so a larger capital to sustain this would be needed.

I'm looking to fund researchers motivated by obsession for knowledge and genuine curiosity, uninterested in social status. Satoshi Nakamoto is a good example of the spirit which I look for: anonymously publishing ground breaking research without looking for recognition. Please reach out to lorenzo.pieri.research[at]gmail.com if you want to chat.

I've pursued personally-funded research for more than a year. Some thoughts:

Think about your competitive advantages. What makes working at your place different from a university and why would a researcher choose yours? It could be they don't like to teach. It could be that they want their life taken care off and you can hire a cook/cleaner/etc. for them. It could be that they don't like writing grants. Or they have trouble winning grants. (E.g., their topic is out of favor for various reasons.) It could be the freedom to work on anything. It could be the location. It could be that they could be the most senior person at your institute. Perhaps they want to start something new. It could be that they love that you think their research is valuable.

Your own institute has weaknesses, but they can be ameliorated. Make it easy to travel to. Putting it near a major university may help earn academic traffic and provide for collaboration.

One way to structure the institute is to have one good mind and many followers on. That is easy to bootstrap, once you get the good mind. The other sustainable structure is to pick a topic and have the largest most-focused group working on that topic. People will come for that, but getting it off the ground will be difficult. Perhaps you can a few good minds right after they get their PhDs and build around that team.

If the key is the people, I suggest you talk to the people. Find out what top researchers don't like about university life.

If you need a topic to focus on, I did research on formal proofs --- writing a mathematical proof on a computer. Every other industry has moved to digital, computer-aided design, except the field most suited to it! The way that mathematicians advance their careers in academia is working against the adoption of the technology. A place outside academia might make more progress in the area.

I worked for a small think tank for a few years recently, there are tonnes of social science and policy outlets out there. It effectively operated as a social enterprise, it did a load of survey work for institutional clients (e.g. UN agencies) that paid the bills whilst the staff worked on their priorities on the side.

The way I'd approach the problem would be to set up a non profit or social enterprise, seek to build a core business and grant portfolio to keep the lights on and work from there.

From another perspective, I'm not entirely sure how OpenAI do things, but there may be insights for you in their approach.

You do really need to be thinking long term, you can either burn your own cash for a couple of years and quit with little to show, or build something durable over a period of decades. Reputation isn't something that appears overnight.

Germany seems a tough place to do it though. You'll need someone with a good head for paperwork.

I pursued something similar. I created the Tree Notation Lab in Oahu (https://lab.treenotation.org/), secured a physical home for it at a university, got funding commitments, started recruiting, and came close to pulling the trigger, but slowed down at the last moment circa January 2020.

The reason I hit the brakes were some specific incidents that made me realize how much I was underestimating how difficult recruiting, retaining, and enabling the level of talent I hoped for. I was probably underestimating those challenges by 2 orders of magnitude.

The majority of folks you'd want to recruit probably have great careers already and probably families, and are likely world class, so in today's market you'd need probably on the order of a $100M - $1B+ war-chest to be able to efficiently compete for them and not have recruitment and retainment as an urgent problem that would keep you up every night.

I instead pivoted and decided to make it a remote first, part time thing. I have a number of advisors and collaborators. New people pitch me ideas and I fund them or not. I sometimes put a few weeks in, sometimes just a day a month. I can scale it organically and pause it when I want. I can be an active part of other remote research groups as well. I made this decision pre-COVID, when it was more of a tossup, but now I'd say it's a no brainer.

If you're just looking to house a group of like minded people perhaps something like https://www.together.casa/ or https://www.pandopooling.com/ might be interesting to check out.

Happy to chat more too if that would be helpful. I definitely encourage you to go for it, as I think we need more of these, but at least for me the better bet was to start remote.

I've tried to do this, with partners and failed. Finding all the people with all the skills you need to get a prototype finished, LET ALONE a field-deployable-serviceable prototype at any level of "interesting scale" is absurdly complex and multi-facted.

In the US most research is targeted as selling things to the military. You might be better in Australia or any other country which offers more social funding for these types of things.

This thread has a lot of good ideas, I wish I'd seen it 5 years ago.

My current project might be useful for you: http://github.com/elasticdotventures/_b00t_

I have no direct experience but I have spent a lot of time thinking about how to realize my own dream of having an artist's residency -- which is actually a similar kind of thing, but temporary. I too am about to acquire the real estate, but don't have the rest set up yet.

One unsolved problem I keep returning to is that the limits of the offer will affect who will come join you. Unless you can give people their dream lab and a ton of money and publication then some of the people you want, will not be able to come even if they really do share your vision.

I think this is an inevitable compromise, but one that should be made very deliberately. Consider your dream researcher. Does she have a family? Can you get them to Germany on the appropriate visas with the mandatory health insurance? Will she have to quit her previous job? Can she afford to go without income for the duration of her research? If not, how much money does she need? What can her husband/wife do there? What if they have a child with special needs? Can they bring their three dogs? Can they be there three days a week? How will they get to Berlin?

For my project, I plan to start very small, since I don't really know what I'm doing and don't have too much money nor time to dedicate to it yet. You might consider doing that too: make the place really nice and get some people there for a shorter period of time, and figure out your growth strategy after you've gone through the mini version with some researchers, and have properly digested their feedback.

And put massive effort (and money assuming you have it) into promoting them and their work: the more it looks like Forschungszentrum Akhann is good for one's career, the more you'll have your choice of researchers.

That, and take care of your liability situation. It's not so bad in Germany as in the US, but consider the possiblity someone gets sucked into the flux capacitor while helping you with your research and it's kind of your fault. You don't want to lose the whole institute over it, nor the personal fortune that has let you set it up. Definitely find a good lawyer with relevant experience.

Good luck with it! I think the world needs more stuff like this.

I'm trying to build something like this myself. I acquired a huge building and am trying to set it up as a science museum, makerspace and business incubator. There will be on-site housing as well.

I'm calling it Serendipic.org.

I haven't had too much luck so far, but I've been working on it for a year and have had nothing but problems with zoning. I can't even get power turned on. They just don't understand what I'm trying to do, and can't comprehend that I don't have every detail of the entire project mapped out in minute detail yet.

Do you know Marek Rosa? https://www.goodai.com/about/

His research grants and building in Prague have this same flavor.

I have a recent experience in exactly this and can connect you to folks from the RAND corporation who just founded a spinoff research think tank. Reach out to me via my email on my profile

I don't know anything about it, but being just one island over I've heard of the Pacific Science Institute, which may be something like you are trying to do ( https://www.pacificscienceinstitute.org/)

There's also some great online research communities, like the https://futureofcoding.org/ slack.

Have you considered using the facilities and proximity to academic infrastructure to host academic conferences instead of full-time research? That might be a more realistic way to attract top talent when you are first starting. You would obviously want to do this in discussion with relevant nearby academic institutions, but being able to offer free accommodations for attendees + maybe some travel expenses could help to lure people for a week event, and a conference with the right attendees coming together can definitely benefit academic research.

A similar option would be to host classes or seminars on cutting edge research, taught by one or a few top researchers and attended by post-docs. Not full-time, but a way to start.

If researchers find it beneficial, if it's a conducive environment and the financials are right, if you master the role of "host," then you might be able to increase the number of conferences/seminars per year and, eventually, with a high-enough frequency, add full-time support staff and longer-term functionality (e.g., researcher-in-residence) that can lead to year-round activity.

Maybe this is a slow approach, but even one well-run conference can be beneficial and you'll be able to learn how to do this and make connections along the way.

I've been exploring the landscape of new types of research institutes/organizations, with a catalog of my findings here: https://arbesman.net/overedge/ I'm happy to discuss the possibilities for research orgs, so please feel free to reach out.

Where are you located in germany?

You might want to take a look at the Shuttleworth Foundation, the Forrest Research Foundation, and the Bill & Melinda Gates Foundations as templates.

They're all billionaire-funded foundations enabling research. Shuttleworth works outside traditional academia, the other two fund research in traditional pathways.

You can consider setting up a joint centre with a university, it would help you inherit their prestige and access students/researchers and them getting funding. I work in a centre that has this strong companies/university collaboration and it's a model that can work well.

> End goal as of now: assemble a few great minds, give them the space to work rent-free in a region with good academic infrastructure, ideally let them collaborate, watch the beauty of it.

My advice is not to expect others to do the work for rent and a “stipend”.

This isn’t a game. AI done wrong will be the end of us.

You will need administrators, the people you are going to bring in are unlilely to have the experience to manage the institutional financial aspects of grants. Also most funding agencies will not give easily to places that are not recognized somehow (nor will reviewers).

I also look for such initiatives. We need something like the Invisible College [1]. And artificial life deserves more than a marginal interest, because is the gate towards understanding biological life as a computation.

[1] https://en.wikipedia.org/wiki/Invisible_College

You might find https://newscience.org/ interesting and relevant.

The founder, Alexey Guzey, is a very interesting person https://twitter.com/alexeyguzey

Just build a startup incubator instead for startups with really interesting early stage ideas.

I've never started a research institute, but I've wanted to for a while. I figure if I were to ever win the lottery I'd start a research institute with the proceeds. Good luck!

Thanks for starting this discussion! I’m starting a small research foundation for studying the biology of aging. I’d be happy to keep in touch!

i did my postgrad at a research institute founded by a well known professor in his field. in the first instance you have to decide what research field/theme you want to target or are passionate about, then get lure leading academics in the field (scout them by published works or industry talks). feel free to get in touch if you want me to connect you.

Have you set up some sort of endowment? How are you ensuring long-term financial viability?

Very interesting idea. Keep us posted.

Its probably very specific to your country.

FTL travel? Is anyone looking into this?