Published on September 16, 2021 11:56 PM GMT

This is a submission to the creative writing contest. See notes at the bottom.

Notes

I made this somewhat off the cuff (for context, it's hastily formatted in Google Slides), but I figured posting it sooner and then possibly modifying it would be better than holding on to it in the futile hope that I would have time to edit it later.

As a result, I would be very grateful for feedback, and would be more than happy to modify this in response to any specific suggestions. I can also imagine adding some more "episodes" (pairs of pages, really). And the illustrations can also be easily modified.

I can send anyone who wants it a pdf version of this.

This is obviously in "picture book" style, but my hope is that it's not exclusively interesting to children.

For convenience, the text of the story below:

“Black bird,” said the blue bird. “Why are you flying so far with that piece of bread? I understand if you're full, but there are birds in our tree who are hungry. Don't you love them?”

“Blue bird,” answered the black bird. “I do love them. But there are more birds in the maple tree, and they are hungrier. And they're smaller birds, too, so a small piece of bread will mean more to them than to a bird in our tree.”

“Black bird!” squawked the blue bird. “What are you doing over there pecking at this oak sprout! Don't you see the man with an axe coming for our tree?”

“Blue bird,” answered the black bird. "I see that man with an axe, and he seems terrible... ...But I also see all our friends swarming and pecking at him. I don't know if he'll get to our tree, but I’m not sure that I would help much by flying over there and pecking at him, too. I do know that this sprout here could grow into a giant oak, displace our home tree, and block our sun. Soon it will be too difficult to stop it from growing too large, so I need to peck at it now.”

“Black bird,” cooed the blue bird. “You're not so bad.”

“Thank you, blue bird,” said the black bird, smiling.

“Although I still can't understand why you won't eat beetles. You can't think they're as important as birds.”

“Well,” said the black bird. “I can't be certain they’re important, but they seem to have interesting lives, and I can eat bread, so it's not that hard to not eat beetles, and...”

“Shush,” said the blue bird. “Let's focus on one thing at a time.”

Thanks for reading, and for encouraging fiction!

Published on September 15, 2021 7:13 AM GMT

Our U.S. COVID-19 response was based on a bet

In April 2020, we launched Project 100, a U.S. COVID-19 response, in partnership with Propel and Stand for Children. To date, the program has delivered one-time $1,000 payments to over 178K Americans living in poverty, with the most recent round going out last week.

For the past decade, our core mission has been to reach people living in extreme poverty. While many millions in the United States are in poverty, they’re typically not facing extreme poverty as it is officially defined (living below $1.90/day).

We decided to launch the program given:

the growing need we saw in the U.S. that could be met with cash

the number of funders interested in getting cash to people in the U.S. who would not otherwise give internationally

the opportunity to raise direct giving’s public profile beyond what we could do with our international programs alone

Our bet, based on points 2 and 3, was that a justified U.S. cash program would also end up drawing in more funding for people living below the extreme poverty line internationally.

We kept funds & attention going to our global work

We continued to focus on driving attention to our global work and took precautions not to divert funds when setting up Project 100:

We only allocated funds specifically donated to the U.S. program to American recipients. No other funding went to the program

We tested and tracked whether donors who first gave to the U.S. would be interested in giving internationally with their future gifts

We maximized press/social media attention on direct giving and ensured the press mentioned our international work wherever possible

Our U.S. work helped make 2021 our best fundraising year yet for international recipients

A year and a half on, the results suggest the bet has paid off. This has already been our strongest year raising funds for people living in poverty internationally with $138.8M YTD in 2021 versus $121.4M and $47.2M total raised for int’l recipients in 2020 and 2019, respectively. Beyond that:

We’ve driven over $70M to international programs from donors who initially gave to U.S. projects — more than our revenue for any year before 2020

Our U.S. program had over 100 press mentions, helping to raise direct cash giving’s profile on the national and global stage

We expect to reach even more international recipients by the end this year than last:

We will continue to work in the U.S.

We expect there to continue to be a legitimate need in the US that can be met with cash, be it after a catastrophic hurricane or in geographies that are chronically poor & under-resourced. We also know there will always be funders who are more focused on giving within the U.S. (perhaps some immutably, and some not). So, moving forward, we’ll launch U.S. programs that we think could both fill a real need and raise direct giving’s profile to drive more support to international recipients.

This was a bet that we’re glad paid off. We’re proud to have helped 178K Americans living in poverty:

Numbers as of Sept 13th, 2021. To hear more from recipients themselves, check out stories from Project 100 recipients and GDLive to hear from folks we’ve reached internationally.

Published on September 15, 2021 7:13 AM GMT

Our U.S. COVID-19 response was based on a bet

In April 2020, we launched Project 100, a U.S. COVID-19 response, in partnership with Propel and Stand for Children. To date, the program has delivered one-time $1,000 payments to over 178K Americans living in poverty, with the most recent round going out last week.

For the past decade, our core mission has been to reach people living in extreme poverty. While many millions in the United States are in poverty, they’re typically not facing extreme poverty as it is officially defined (living below $1.90/day).

We decided to launch the program given:

the growing need we saw in the U.S. that could be met with cash

the number of funders interested in getting cash to people in the U.S. who would not otherwise give internationally

the opportunity to raise direct giving’s public profile beyond what we could do with our international programs alone

Our bet, based on points 2 and 3, was that a justified U.S. cash program would also end up drawing in more funding for people living below the extreme poverty line internationally.

We kept funds & attention going to our global work

We continued to focus on driving attention to our global work and took precautions not to divert funds when setting up Project 100:

We only allocated funds specifically donated to the U.S. program to American recipients. No other funding went to the program

We tested and tracked whether donors who first gave to the U.S. would be interested in giving internationally with their future gifts

We maximized press/social media attention on direct giving and ensured the press mentioned our international work wherever possible

Our U.S. work helped make 2021 our best fundraising year yet for international recipients

A year and a half on, the results suggest the bet has paid off. This has already been our strongest year raising funds for people living in poverty internationally with $138.8M YTD in 2021 versus $121.4M and $47.2M total raised for int’l recipients in 2020 and 2019, respectively. Beyond that:

We’ve driven over $70M to international programs from donors who initially gave to U.S. projects — more than our revenue for any year before 2020

Our U.S. program had over 100 press mentions, helping to raise direct cash giving’s profile on the national and global stage

We expect to reach even more international recipients by the end this year than last:

We will continue to work in the U.S.

We expect there to continue to be a legitimate need in the US that can be met with cash, be it after a catastrophic hurricane or in geographies that are chronically poor & under-resourced. We also know there will always be funders who are more focused on giving within the U.S. (perhaps some immutably, and some not). So, moving forward, we’ll launch U.S. programs that we think could both fill a real need and raise direct giving’s profile to drive more support to international recipients.

This was a bet that we’re glad paid off. We’re proud to have helped 178K Americans living in poverty:

Numbers as of Sept 13th, 2021. To hear more from recipients themselves, check out stories from Project 100 recipients and GDLive to hear from folks we’ve reached internationally.

Published on September 10, 2021 7:47 AM GMT

On April 21 the European Commission announced their proposal for European regulation of AIThe proposal strives to be for AI what GDPR has been for data protection.Over the coming years the proposal will go through multiple readings in the Parliament and Council, where modifications can be proposed to the act before it is finally adopted or dismissed.

After the AI Act was announced, most attention fell on the act's wide definition of AI and blanket bans on 'manipulative AI'. Unfortunately this focus has led many to miss out on the most important points of the act.

I noticed there were no forum posts that would help someone get up to speed on the act. In this post I will summarise the act's most important points, how it may affect the development of transformative AI, as well as the EA community's response to the proposal.

What are the act's important points?

I below outline what I deem the most important points of the regulation, based on their effect on the development of transformative AI. I skip lightly over many details of the act, such as regulatory sandboxes and special rules for biometric systems, to keep the summary brief.

The act will apply to all EU countries and supersede any conflicting national law. Because of the act's broad definition of AI, it is difficult for any EU countries to make laws on AI which would not be in conflict.

It does not apply to military use of AI. Here countries are free to do as they see fit.

Rules for 'high-risk' AI

The Act lists a series of 'high-risk' areas. Systems operating in a high-risk area are considered high-risk and must be reviewed and approved before they can be placed on the market. This means that the AI Act's regulation will not apply to AI developed and used internally by companies. High-risk systems include everything from AI management of electricity grids, to AI that determines who to promote or fire.

Areas that are considered high-risk where certain uses are restricted are the following:

biometric identification and categorization

management and operation of critical infrastructure

educational and vocational training

employment, worker management, and access to self-employment

access to and enjoyment of essential services and benefits

law enforcement

migration, asylum and border management

administration of justice and democracy

A red thread across the systems which are considered high-risk, is that they make decisions which significantly affect the lives of citizens. The full list of high-risk systems is two pages and can be read in Annex III.

After the law is passed, the commission can add new uses of AI that must be approved as long as they fall under any of the existing high-risk areas.

For a high-risk system to be approved the provider must submit detailed technical documentation for the system.¹ Requirements for technical documentation include:

design specifications, key design choices, description of what the system is optimizing.

description of any use of third party tools.

description of training data, how it has been obtained, how it has been processed.

How the system can be monitored and controlled.

The system must have in-built operational constraints that cannot be overridden by the system itself and is responsive to the human operator (!!)

Description of foreseeable risks the system poses to EU citizens' health, safety and fundamental rights.

In other words the act creates a large, (partially) updatable list of areas where nobody is allowed to deploy AI without explicitly getting approval, which you only get by living up to numerous safety requirements of which one is a working off-switch. High-risk systems not only need approval, but must also be continuously monitored after they are placed on the market.

Establishment of the European AI Board

To enforce this, the act requires new institutions to be created.

The European AI Board.

Run by the EU Commission (EU's civil service).

Oversees national authorities and...

Published on September 10, 2021 7:47 AM GMT

On April 21 the European Commission announced their proposal for European regulation of AIThe proposal strives to be for AI what GDPR has been for data protection.Over the coming years the proposal will go through multiple readings in the Parliament and Council, where modifications can be proposed to the act before it is finally adopted or dismissed.

After the AI Act was announced, most attention fell on the act's wide definition of AI and blanket bans on 'manipulative AI'. Unfortunately this focus has led many to miss out on the most important points of the act.

I noticed there were no forum posts that would help someone get up to speed on the act. In this post I will summarise the act's most important points, how it may affect the development of transformative AI, as well as the EA community's response to the proposal.

What are the act's important points?

I below outline what I deem the most important points of the regulation, based on their effect on the development of transformative AI. I skip lightly over many details of the act, such as regulatory sandboxes and special rules for biometric systems, to keep the summary brief.

The act will apply to all EU countries and supersede any conflicting national law. Because of the act's broad definition of AI, it is difficult for any EU countries to make laws on AI which would not be in conflict.

It does not apply to military use of AI. Here countries are free to do as they see fit.

Rules for 'high-risk' AI

The Act lists a series of 'high-risk' areas. Systems operating in a high-risk area are considered high-risk and must be reviewed and approved before they can be placed on the market. This means that the AI Act's regulation will not apply to AI developed and used internally by companies. High-risk systems include everything from AI management of electricity grids, to AI that determines who to promote or fire.

Areas that are considered high-risk where certain uses are restricted are the following:

biometric identification and categorization

management and operation of critical infrastructure

educational and vocational training

employment, worker management, and access to self-employment

access to and enjoyment of essential services and benefits

law enforcement

migration, asylum and border management

administration of justice and democracy

A red thread across the systems which are considered high-risk, is that they make decisions which significantly affect the lives of citizens. The full list of high-risk systems is two pages and can be read in Annex III.

After the law is passed, the commission can add new uses of AI that must be approved as long as they fall under any of the existing high-risk areas.

For a high-risk system to be approved the provider must submit detailed technical documentation for the system.¹ Requirements for technical documentation include:

design specifications, key design choices, description of what the system is optimizing.

description of any use of third party tools.

description of training data, how it has been obtained, how it has been processed.

How the system can be monitored and controlled.

The system must have in-built operational constraints that cannot be overridden by the system itself and is responsive to the human operator (!!)

Description of foreseeable risks the system poses to EU citizens' health, safety and fundamental rights.

In other words the act creates a large, (partially) updatable list of areas where nobody is allowed to deploy AI without explicitly getting approval, which you only get by living up to numerous safety requirements of which one is a working off-switch. High-risk systems not only need approval, but must also be continuously monitored after they are placed on the market.

Establishment of the European AI Board

To enforce this, the act requires new institutions to be created.

The European AI Board.

Run by the EU Commission (EU's civil service).

Oversees national authorities and...

Published on September 15, 2021 2:52 PM GMT

Melanie Perkins, the CEO of Canva, recently announced that they intend to donate the “vast majority of their equity to do good in the world”. This comes out to at least $6 billion. I think this is awesome news and a potentially exciting opportunity for the EA community.

As reported in several media outlets, Canva has recently been valued at $40 billion USD. Following this, Melanie Perkins (Canva’s CEO) outlined in a blog post their vision, through a simple Two-Step plan:

With a $40 billion valuation, Canva has made a lot of progress on step 1. To achieve step 2, Melanie and her husband and co-founder, Cliff, have “committed the vast majority of their equity (30% of Canva) to do good in the world”, through the Canva Foundation.[1] This works out to be the equivalent of at least $6 billion USD, and potentially more.[2] To put this into perspective, that would be more than 10 times that of all Giving What We Can pledges (using Ben Todd’s estimates).

Canva have already taken the 1% pledge,[3] but are now looking to expand their philanthropy further. Excitingly for EAs, this starts with a $10 million pilot program with GiveDirectly in South Africa. They then “hope to rapidly scale this more broadly and to contribute to the lives of as many people across the globe as we can.” [emphasis added]

This strikes me as both awesome news and an exciting opportunity. What could/should EAs do about this? I have some very tentative suggestions:

Find out more. I heard about this today, and have only spent ~1 hour Googling. Investigating more about this news and the Canva Foundation seems potentially worthwhile, to get a better grip on what this could mean for EAs.

Maybe nothing. There are potential backfire risks when doing outreach, the CEO may have their own priorities, or perhaps the money is already committed to EA causes, so action is less valuable.

Something else? I have spent very little time looking into this, but would be interested in other people's thoughts.

Disclaimer: I have no associations or affiliations with Canva.

Published on September 15, 2021 2:52 PM GMT

Melanie Perkins, the CEO of Canva, recently announced that they intend to donate the “vast majority of their equity to do good in the world”. This comes out to at least $6 billion. I think this is awesome news and a potentially exciting opportunity for the EA community.

As reported in several media outlets, Canva has recently been valued at $40 billion USD. Following this, Melanie Perkins (Canva’s CEO) outlined in a blog post their vision, through a simple Two-Step plan:

With a $40 billion valuation, Canva has made a lot of progress on step 1. To achieve step 2, Melanie and her husband and co-founder, Cliff, have “committed the vast majority of their equity (30% of Canva) to do good in the world”, through the Canva Foundation.[1] This works out to be the equivalent of at least $6 billion USD, and potentially more.[2] To put this into perspective, that would be more than 10 times that of all Giving What We Can pledges (using Ben Todd’s estimates).

Canva have already taken the 1% pledge,[3] but are now looking to expand their philanthropy further. Excitingly for EAs, this starts with a $10 million pilot program with GiveDirectly in South Africa. They then “hope to rapidly scale this more broadly and to contribute to the lives of as many people across the globe as we can.” [emphasis added]

This strikes me as both awesome news and an exciting opportunity. What could/should EAs do about this? I have some very tentative suggestions:

Find out more. I heard about this today, and have only spent ~1 hour Googling. Investigating more about this news and the Canva Foundation seems potentially worthwhile, to get a better grip on what this could mean for EAs.

Maybe nothing. There are potential backfire risks when doing outreach, the CEO may have their own priorities, or perhaps the money is already committed to EA causes, so action is less valuable.

Something else? I have spent very little time looking into this, but would be interested in other people's thoughts.

Disclaimer: I have no associations or affiliations with Canva.

Published on September 15, 2021 12:32 PM GMT

In a 2021 MotherJones article, Sinduja Rangarajan, Tom Philpott, Allison Esperanza, and Alexis Madrigal compiled and visualized 186 publicly available predictions about timelines for culture meat (made primarily by cultured meat companies and a handful of researchers). I added 11 additional predictions ACE had collected, and 76 other predictions I found in the course of a forthcoming Rethink Priorities project.

Of the 273 predictions collected, 84 have resolved - nine resolving correctly, and 75 resolving incorrectly. Additionally, another 40 predictions should resolve at the end of the year and look to be resolving incorrectly. Overall, the state of these predictions suggest very systematic overconfidence. Cultured meat seems to have been perpetually just a few years away since as early as 2010 and this track record plausibly should make us skeptical of future claims from producers that cultured meat is just a few years away.

Here I am presenting the results of predictions that have resolved, keeping in mind they are probably not a representative sample of publicly available predictions, nor assembled from a systematic search. Many of these are so vaguely worded that it’s difficult to resolve them positively or negatively with high confidence. Few offer confidence ratings, so we can’t measure calibration.

Below is the graphic made in the MotherJones article. It is interactive in the original article.

The first sale of a ~70% cultured meat chicken nugget occured in a restaurant in Singapore on 2020 December 19th for S$23 (~$17 USD), created by Eat Just at a loss to the company.

65 predictions made on cultured meat being available on the market or in supermarkets specifically can now be resolved. 56 were resolved negatively and in the same direction - overly optimistic (update: the original post said 52). None resolved negatively for being overly pessimistic. These could resolve differently depending on your exact interpretation but I don't think there is an order of magnitude difference in interpretations. The nine that plausibly resolved positively are listed below.

In 2010 "At least another five to 10 years will pass, scientists say, before anything like it will be available for public consumption". (A literal reading of this resolves correct, even though one might interpret the meaning as a product will be available soon after ten years)

Mark Post of Maastricht University & Mosa Meat in 2014 stated he “believes a commercially viable cultured meat product is achievable within seven years." (It’s debatable if the Eat Just nugget is commercially viable as it is understood to be sold at a loss for the company).

Peter Verstate of Mosa Meat in 2016 predicted that premium priced cultured products should be available in 5 years (ACE 2017)

Mark Post in 2017 "says he is happy with his product, but is at least three years from selling one" (A literal reading of this resolves correct, even though one might interpret the meaning as a product will be available soon after three years)

Bruce Friedrich of the Good Food Institute in March 2018 predicted “clean-meat products will be available at a high price within two to three years”

Unnamed scientists in December 2018 “say that you can buy it [meat in a laboratory from cultured cells] within two years."

Aleph Farms in December 2018 saying their “product will not be commercially available for at least two years." (A literal reading of this resolves correct, even though one might interpret the meaning as a product will be available soon after two years)

Kate Krueger, PhD, a cell biologist and the director of research for New Harvest said in January 2020"We're still probably a decade away from lab-grown hot wings, but cultured chicken nuggets and burgers might be available in the next two years"

Eat Just claimed they would sell the chicken nugget that week in December 2020

Note that the companies above also made additional predictions that were i...

Published on September 15, 2021 12:32 PM GMT

In a 2021 MotherJones article, Sinduja Rangarajan, Tom Philpott, Allison Esperanza, and Alexis Madrigal compiled and visualized 186 publicly available predictions about timelines for culture meat (made primarily by cultured meat companies and a handful of researchers). I added 11 additional predictions ACE had collected, and 76 other predictions I found in the course of a forthcoming Rethink Priorities project.

Of the 273 predictions collected, 84 have resolved - nine resolving correctly, and 75 resolving incorrectly. Additionally, another 40 predictions should resolve at the end of the year and look to be resolving incorrectly. Overall, the state of these predictions suggest very systematic overconfidence. Cultured meat seems to have been perpetually just a few years away since as early as 2010 and this track record plausibly should make us skeptical of future claims from producers that cultured meat is just a few years away.

Here I am presenting the results of predictions that have resolved, keeping in mind they are probably not a representative sample of publicly available predictions, nor assembled from a systematic search. Many of these are so vaguely worded that it’s difficult to resolve them positively or negatively with high confidence. Few offer confidence ratings, so we can’t measure calibration.

Below is the graphic made in the MotherJones article. It is interactive in the original article.

The first sale of a ~70% cultured meat chicken nugget occured in a restaurant in Singapore on 2020 December 19th for S$23 (~$17 USD), created by Eat Just at a loss to the company.

65 predictions made on cultured meat being available on the market or in supermarkets specifically can now be resolved. 56 were resolved negatively and in the same direction - overly optimistic (update: the original post said 52). None resolved negatively for being overly pessimistic. These could resolve differently depending on your exact interpretation but I don't think there is an order of magnitude difference in interpretations. The nine that plausibly resolved positively are listed below.

In 2010 "At least another five to 10 years will pass, scientists say, before anything like it will be available for public consumption". (A literal reading of this resolves correct, even though one might interpret the meaning as a product will be available soon after ten years)

Mark Post of Maastricht University & Mosa Meat in 2014 stated he “believes a commercially viable cultured meat product is achievable within seven years." (It’s debatable if the Eat Just nugget is commercially viable as it is understood to be sold at a loss for the company).

Peter Verstate of Mosa Meat in 2016 predicted that premium priced cultured products should be available in 5 years (ACE 2017)

Mark Post in 2017 "says he is happy with his product, but is at least three years from selling one" (A literal reading of this resolves correct, even though one might interpret the meaning as a product will be available soon after three years)

Bruce Friedrich of the Good Food Institute in March 2018 predicted “clean-meat products will be available at a high price within two to three years”

Unnamed scientists in December 2018 “say that you can buy it [meat in a laboratory from cultured cells] within two years."

Aleph Farms in December 2018 saying their “product will not be commercially available for at least two years." (A literal reading of this resolves correct, even though one might interpret the meaning as a product will be available soon after two years)

Kate Krueger, PhD, a cell biologist and the director of research for New Harvest said in January 2020"We're still probably a decade away from lab-grown hot wings, but cultured chicken nuggets and burgers might be available in the next two years"

Eat Just claimed they would sell the chicken nugget that week in December 2020

Note that the companies above also made additional predictions that were i...

Published on September 14, 2021 8:43 PM GMT

Epistemic status: Half-baked at best

I have often been skeptical of the value of a) critiques against effective altruism and b) fully general arguments that seem like they can apply to almost anything. However, as I am also a staunch defender of hypocrisy, I will now hypocritically attempt to make the case for applying a fully general critique to effective altruism.

In this post, I will claim that:

Motivated reasoning inhibits our ability to acquire knowledge and form reasoned opinions.

Effective altruism should not be assumed to be above these biases. Moreover, there are strong reasons to believe that incentive structures and institutions in effective altruism exacerbate rather than alleviate these biases.

Observed data and experiences in effective altruism support this theory; they are consistent with an environment where motivated reasoning and selection biases are rampant.

To the extent that these biases (related to motivated reasoning) are real, we should expect the harm done to our ability to form reasoned opinions to also seriously harm the project of doing good.

I will use the example of cost-effectiveness analyses as a jumping board for this argument. (I understand that effective altruism, especially outside of global health and development, has largely moved away from explicit expected value calculations and cost-effectiveness analyses. However, I do not believe this change invalidates my argument (see Appendix B)).

I also list a number of tentative ways to counteract motivated reasoning and selection bias in effective altruism:

Encourage and train scientific/general skepticism in EA newcomers.

Try marginally harder to accept newcomers, particularly altruistically motivated ones with extremely high epistemic standards

As a community, fund and socially support external (critical) cost-effectiveness analyses and impact assessments of EA orgs

Within EA orgs, encourage and reward dissent of various forms

Commit to individual rationality and attempts to reduce motivated reasoning

Maybe encourage a greater number of people to apply and seriously consider jobs outside of EA or EA-adjacent orgs

Maintain or improve the current culture of relatively open, frequent, and vigorous debate

Foster a bias towards having open, public discussions of important concepts, strategies, and intellectual advances

Motivated reasoning: What it is, why it’s common, why it matters

By motivated reasoning, I roughly mean what Julia Galef calls “soldier mindset” (H/T Rob Bensinger):

In directionally motivated reasoning, often shortened to "motivated reasoning", we disproportionately put our effort into finding evidence/reasons that support what we wish were true.

Or, from Wikipedia:

emotionally biased reasoning to produce justifications or make decisions that are most desired rather than those that accurately reflect the evidence

I think motivated reasoning is really common in our world. As I said in a recent comment:

My impression is that my interactions with approximately every entity that perceives themself as directly doing good outside of EA is that they are not seeking truth, and this systematically corrupts them in important ways. Non-random examples that come to mind include public health (on covid, vaping, nutrition), bioethics, social psychology, developmental econ, climate change, vegan advocacy, religion, US Democratic party, and diversity/inclusion. Moreover, these problems aren't limited to particular institutions: these problems are instantiated in academia, activist groups, media, regulatory groups and "mission-oriented" companies.

What does motivated reasoning look like in practice? In the field of cost-effectiveness analyses, it might look like this comment on a blog post about scientific conflicts of interest:

Back in the 90’s I did some consulting work for a startup that was developing a new medical device. They were honest people–they never pressured me. My contract stipulated that ...