Balsa

Wow, what a year it has been. Things keep getting crazier.

Thank you for taking this journey with me. I hope I have helped you keep pace, and that you have been able to discern for yourself the parts of this avalanche of words and events that were helpful. I hope to have helped things make somewhat more sense.

And I hope many of you have taken that information, and used it not only to be able to check Twitter less, but also to make better decisions, and, hopefully, to help make the world a better place—one in which humanity is more likely to survive.

Recently, my coverage of the Biden administration executive order and  the events at OpenAI have been received very positively. I’d like to do more in that mold: more focused, shorter pieces that pull the story together, hopefully de-emphasizing more ephemeral weekly posts over time. I am also happy that this work has potentially opened doors that might grant me larger platforms and other ways to make a difference.

If you feel it would make the world better to do so, please help spread the word to others who would find my work useful.

Thank you especially to  both my long-time and recent paid subscribers and my Patreon supporters. It is important to me that all my content remain freely accessible— so please do not subscribe if it would be a hardship—but subscriptions and other contributions are highly motivating and allow me to increase my budget.

You can also help by contributing to my 501c(3), Balsa Research. We need your help in order to continue our work.

The rest of this post is an update on what is happening there.

Even with the craziness that is AI, it is important not to lose sight of everything else going on and to seek out opportunities to create a better, saner world. That means building a world that’s better equipped to handle AI’s challenges and one that knows it can do sensible things.

Previously I shared both an initial announcement and Balsa FAQ. Since then, we’ve focused on identifying particularly low-hanging fruit where key bridging work is not being done. I’ve hired Jennifer Chen, who has set up the necessary legal and logistical infrastructure for us to begin work. We’ve had a lot of conversations with volunteers, considered many options and game plans, and are ready to begin work in earnest.

As our first major project, we’ve decided that Balsa will work to repeal the Jones Act. That is a big swing, and we are small. We feel that the current approaches in Jones Act reform are flawed and that there’s an opportunity here to really move the needle (but, if it turns out we’re wrong, we will pivot).

Our plan continues to be to lay the necessary groundwork for a future push.

We’ll prioritize identifying the right questions to ask and commissioning credible academic work to find and quantify those answers. The questions that matter are often going to be the ones that are going to matter in a Congressional staff meeting or hearing, breaking down questions that particular constituencies and members care about.

I believe that the numbers will show both big wins and few net losers from repeal—including few losers among the dedicated interest groups that are fighting for the Jones Act tooth and nail—such that full compensation for those losers would be practical. We also think that the framing and understanding of the questions involved can be dramatically improved. The hope is that the core opposition, which comes largely from unions, can ultimately be brought into a win-win deal.

This is somewhat of a narrowing of the mission. The intended tech arm of Balsa did not end up happening. We will not attempt to influence elections or support candidates. We do not anticipate having the resources for the complete stack, although, if we get broader support than anticipated, we will explore what that enables.

In the short term, if we secure the funding, we will be able to continue to pay Jennifer, coordinate volunteers and to commission studies—and, later, commission the drafting of legislative language. We’ll then consider how else we might scale and make progress. Ideally, if we get traction, others will help as well.

Balsa also has three other intended cause areas: NEPA, housing reform, and inevitably AI.

Housing is on the list because I believe in the housing theory of everything. It is at the heart of our economic problems and of many people’s impoverished lived experiences.

I believe Balsa’s comparative advantage on housing would be to look at interventions at the federal level, where it seems like options for policy reform are neglected. Progress so far has been remarkable at the local and state level, but we believe there is room to expand upon that at the Federal level.

The first key is that the incentive structure at the federal level is far better-suited for reform. Everyone, even homeowners, appreciate that home prices are too high in the U.S. and that it would be good to build more housing overall—but are too wedded to the status quo to support those efforts if it’d affect their neighborhoods. To the extent we can shift the locus of decision-making from localities to the federal government, we can change the housing debate from a fight over streets and blocks to one where those higher-level incentives start to bind. The deal gets better.

The other key is that Congress and the White House can lean on levers that just aren’t available to a state or city. Federal policy controls home loans and interest rates. It controls a wide variety of related funds including Fannie Mae and Freddie Mac. It controls or could control various standards and rules. The housing market is intertwined nationally, allowing the invocation of the Constitution’s Commerce Clause. If there is a will, we can find a way.

As with the Jones Act, the prize is large. Even though the odds are stacked against us, the attempt is still worthwhile.

This is an even longer shot and bigger swing. Someone needs to enable an attempt.

NEPA is based on the principle that, in order to build something, one should first be required to submit all the necessary paperwork. Then, once that paperwork is filed, others can sue saying it is insufficient or not in order. Then, once the paperwork is ruled to be in order, things can proceed whether or not the underlying project makes sense.

Over time, this has resulted in increasingly absurdist quantities of required paperwork. It takes years to complete a NEPA review whether or not a project raises actual environmental issues.

This is crazy. Rather than patch the worst parts of the law, we should replace the whole thing.

Balsa’s plan is to flesh out an entirely different approach. Instead of requiring paperwork at all, the U.S. should require developers to conduct cost-benefit analysis on the project, then evaluate the results.

The new proposed framework is that, if a full review would be required under NEPA, an independent analysis by a qualified private entity must be commissioned instead. The entity then has a fixed period of time—and a budget that is a function of the ultimate project budget—with which to perform its analysis. During that time, the project can be modified, and those modifications incorporated into the analysis. Compensation to stakeholders or other negotiated deals can be part of the proposal.

Once that analysis is complete, the government composes a panel that includes all major stakeholders. The panel reviews the analysis and votes on whether the project can proceed. As part of the decision, the panel can consider whether the analysis is credible and complete, and those who have objections can raise them.

Under this model there is no set of formal requirements for considerations or paperwork, and there is no mechanism that grants others standing to sue and block the project. The advantages of a development would be given equal weight as its disadvantages, especially for green energy projects.

If, after the vote, certain groups continue to claim that they are entitled to civil remedy because of the damage that would be done or is being done by the project, they can make their case—but even if they win compensation, the project will continue.

That is a rough start of the sketch of a complex proposal that will need to be fleshed out. Operationalizing the details will not be easy, but that’s the point. Someone has to make these ideas concrete so that they can be considered and discussed.

A year ago, when I founded Balsa, I had no intention of getting involved with AI.

We all know how that worked out.

I still think that getting involved in direct AI policy advocacy would be a mistake (Think of  the policy advocacy ecosystem as the xkcd comic with the 15 different standards). It’s much better to fund and guide existing efforts than to start yet another one. I am not about to go down to Washington DC, and I am not about to take a ton of meetings. Let someone else do that.

Am I bracing for the possibility I am once again overtaken by events and feel compelled to take a more active role there? It could happen. If I get a bunch of people who want to dedicate funds to that purpose, I will certainly look into it. Still, I hope to avoid that outcome.

The good news is that, unlike a year ago, we have a good idea what reasonable incremental policy will need to look like, and we are at least somewhat on track to making that happen.

Policy now has to lay the groundwork for a regime where we have visibility into the training of frontier models. We collectively must gain the power to restrict such training once a model’s projected capabilities become existentially dangerous until we are confident we know what it would take to proceed safely.

That means registration of large data centers and concentrations of compute. That means, at minimum, registration of any training runs above some size threshold, such as the 10^26 flops chosen in the executive order. In the future, it means requiring additional safeguards, up to and including halting until we figure out sufficient additional procedures, with those procedures ramping up alongside projected and potential model capabilities. Those include computer security requirements, safeguards against mundane harm and misuse, and, most importantly, protection against existential risks.

I believe that such a regime will automatically be a de facto ban on sufficiently capable open source frontier models. Alignment of such models is impossible; it can easily be undone. The only way to prevent misuse of an open source model is for the model to lack the necessary underlying capabilities. Securing the weights of such models is obviously impossible by construction. Releasing a sufficiently capable base model now risks releasing the core of a future existential threat when we figure out the right way to scaffold on top of that release—a release that cannot be taken back.

Finally, we will need an international effort to extend such standards everywhere.

Yes, I know there are those who say that this is impossible, that it cannot be done. To them, I say: the alternative is unthinkable, and many similarly impossible things happen when there is no alternative.

Yes, I know there are those who would call such a policy various names, with varying degrees of accuracy. I do not care.

I would also say to such objections that the alternative is worse. Failure to regulate at the model level, even if not directly fatal, would then require regulation at the application level, when the model implies the application for any remotely competent user.

Give every computer on the planet the power to do that which must be policed, and you force the policing of every computer. Let’s prevent the need for that.

There are also many other incrementally good policies worth pursuing. I am happy to help prevent mundane harms and protect mundane utility, and explore additional approaches. This can be a ‘yes, and’ situation.

Again, I believe my place in AI is primarily outside Balsa:in writing, in making connections, in laying the rhetorical groundwork and seeking clarity and understanding. But I have been wrong about such things before.

I want to be clear up front: Thanks to generous supporters who prefer to remain anonymous, my personal financial situation allows me to devote my full time efforts where I believe they can produce the most value without personally worrying about money. Working as best I can, I am close to my production possibilities frontier.

That does not mean I have unlimited funds, or that marginal additional personal funding or subscriptions would not be valuable. Paid subscriptions are highly motivating, expand the budget in useful ways, and are very much appreciated. Sufficient additional resources would allow me to enlist some combination of editing, journalistic, operational and engineering help, and otherwise explore ways to enhance and expand my production and productivity.

Balsa will require funding. We need to pay Jennifer Chen, the person who will handle operations and coordination of volunteers. We will also need a budget to commission academic studies and other work and for other general operations, including drafting legislative language. Until that is securely in hand, I will be taking zero compensation from Balsa beyond expenses. If support is sufficiently generous, we can look to expand.

There is lots to do now that the one-time fixed costs of having an operational 501c(3) have been paid. Balsa needs additional funding in order to take advantage of the opportunities we see. You can support us here.