Tag Archives: Research

AI As Software Grant

While I’ve been part of grants before, and had research support, I’ve never had support for my futurist work, including the years I spent writing Age of Em. That now changes:

The Open Philanthropy Project awarded a grant of $264,525 over three years to Robin Hanson (Associate Professor of Economics, George Mason University) to analyze potential scenarios in the future development of artificial intelligence (AI). Professor Hanson plans to focus on scenarios in which AI is developed through the steady accumulation of individual pieces of software and leads to a “multipolar” outcome. .. This grant falls within our work on potential risks from advanced artificial intelligence, one of our focus areas within global catastrophic risks. (more)

Who is Open Philanthropy? From their summary:

Good Ventures is a philanthropic foundation whose mission is to help humanity thrive. Good Ventures was created by Dustin Moskovitz (co-founder of Facebook and Asana) and Cari Tuna, who have pledged to give the majority of their wealth to charity. .. GiveWell is a nonprofit that finds outstanding giving opportunities and publishes the full details of its analysis to help donors decide where to give. .. The Open Philanthropy Project is a collaboration between Good Ventures and GiveWell in which we identify outstanding giving opportunities, make grants, follow the results, and publish our findings.

A key paragraph from my proposal:

Robin Hanson proposes to take three years to conduct a broad positive analysis of the multipolar scenario wherein AI results from relatively steady accumulation of software tools. That is, he proposes to assume that human level AI will result mainly from the continued accumulation of software tools and packages, with distributions of cost and value correlations similar to those seen so far in software practice, in an environment where no one actor dominates the process of creating or fielding such software. He will attempt a mostly positive analysis of the social consequences of these assumptions, both during and after a transition to a world dominated by AI. While this is hardly the universe of all desired analyses, it does seem to cover a non-trivial fraction of interesting cases.

I and they see value in such an analysis even if AI software ends up differing systematically from the software we’ve seen so far:

While we do not believe that the class of scenarios that Professor Hanson will be analyzing is necessarily the most likely way for future AI development to play out, we expect his research to contribute a significant amount of useful data collection and analysis that might be valuable to our thinking about AI more generally, as well as provide a model for other people to follow when performing similar analyses of other AI scenarios of interest.

My idea is to extract from our decades of experience with software a more detailed description of the basic economics of software production and use. To distinguish, as time allows, many different kinds of inputs to production, styles of production, parts of produced products, and types of uses. And then to sketch out different rough “production functions” appropriate to different cases. That is, to begin to translate basic software engineering insight into economics language.

The simple assumption that software doesn’t fundamentally change in the future is the baseline scenario, to be fed into standard economic models to see what happens when such a more richly described software sector slowly grows to take over the economy. But a richer more detailed description of software economics can also give people a vocabulary for describing their alternative hypotheses about how software will change. And then this analysis framework can be adjusted to explore such alternative hypotheses.

So right from the start I’d like to offer this challenge:

Do you believe that the software that will let machines eventually do pretty much all jobs better than humans (or ems) will differ in foreseeable systematic ways from the software we have seen in the last seventy years of software practice? If so, please express your difference hypothesis as clearly as possible in terminology that would be understandable and familiar to software engineers and/or economists.

I will try to stretch the economic descriptions of software that I develop in the direction of encompassing the most common such hypotheses I find.

GD Star Rating
loading...
Tagged as: , , ,

Middle-aged not miserable, just too busy to answer surveys

I have read several times that there is evidence of a U-curve in happiness over an individual’s life. People are happy in their youth, and happy again after retirement, but suffer from a serious malaise in between as they grapple with their finances, careers and family life. Seeing as how I’m about to embark on this part of my life, that isn’t a particularly appealing idea! Today I was glad to find some evidence that the U-curve is just a statistical illusion. Australian economists Paul Frijters and Tony Beatton have analysed large panel data sets from Australia (n>10,000), the UK (n>25,000) and Germany (n>20,000) and produced the following trajectory for happiness as people age:

Happiness is nearly flat from 20 through 50. Frijter’s explanation for the disagreement with the existing literature is that,

previous studies severely underestimated the degree to which miserable people in middle-age were over-represented in these datasets: happy people in middle age are busy and don’t have time to participate in lengthy questionnaires, leading previous studies to erroneously think there was a huge degree of unhappiness in middle-age. When you actually follow people over time, no such ‘middle-age blues pattern’ can be found, at least not in Australia or the UK and only to a mild degree in Germany. … we found that there were severe data problems in Germany, with only quite miserable people in middle-age prepared to partake in the sample and respondents becoming markedly more honest (and miserable) as they answered the happiness questions year after year.

Those of us with active lives need not fear! There are familiar selection problems later in life as well. Unhappy people do not live as long, and may become less able or willing to answer surveys than their healthy and cheerful counterparts. By tracking the same participants for 10 years or more, Frijters claims to have “dealt with both issues,” presumably suing their initial responses to by forecast their missing responses later in life. Some day I should choose carefully where I decide to retire and perhaps return to my homeland: “life in old age is clearly relatively better in Australia than the UK, perchance because of the better weather, more generous public pensions, and more space.”

Another, more depressing, paper on happiness recently caught my attention. Probably due to its early use of twins to investigate the causes of happiness it has landed over 900 citations, despite a limited sample of separated twins. It finds a remarkably high correlation between the happiness of twins who share all of their genes, but very little correlation for twins who share half of their genes:

Happiness or subjective wellbeing was measured on a birth-record based sample of several thousand middle-aged twins using the Well Being (WB) scale of the Multidimensional Personality Questionnaire (MPQ). Neither socioeconomic status (SES), educational attainment, family income, marital status, nor an indicant of religious commitment could account for more than about 3% of the variance in WB. From 44% to 53% of the variance in WB, however, is associated with genetic variation. Based on the retest of smaller samples of twins after intervals of 4.5 and 10 years, we estimate that the heritability of the stable component of subjective wellbeing approaches 80%.

For the 48 DZ [fraternal] pairs, this cross-twin, cross-time correlation for WB was essentially zero (.07) while, for the 79 MZ [identical] pairs, it equaled .40, or 80% of the retest correlation of .50. The MZ data suggest that the stable component of wellbeing (i.e., trait-happiness) is largely determined genetically. The negligible DZ correlation suggests that this stable and heritable component of happiness is an emergenic trait (Lykken, 1982; Lykken, Bouchard, McGue, & Tellegen, 1992), that is, a trait that is determined by a configural rather than an additive function of components. Emergenic traits, although determined in part genetically, do not tend to run in families as do traits that are polygenic-additive.

A similar result was reported in an earlier study of 217 MZ and 114 DZ pairs of middle-aged Minnesota Registry twins, plus 44 MZ and 27 DZ pairs who were separated in infancy and reared apart (Tellegen, et al., 1988). The best estimate of the heritability of WB in that study was .48 (± .08) and, as was true here, a model involving only additive genetic effects did not fit the data.

Myers and Diener suggested that people who enjoy close personal relationships, who become absorbed in their work, and who set themselves achievable goals and move toward them with determination are happier on the whole than people who do not. We agree, but we question the direction of the causal arrow. We know that when people with bipolar mood disored are depressed, they tend to avoid intimate encounters or new experiences and tend to brood upon depressing thoughts rather than concetrating on their work. Then, when their moods swings toward elation, these same people tend to do the things that happy people do. This is undoubtedly a James-Lange feedback effect: Dysfunctional behavior exacerbates depression, whereas the things happy people do enhance their happiness. We argue, however, that the impetus is greater from mood to behavior than in the reverse direction. It may be that trying to be happier is as futile as trying to be taller and therefore is counterproductive. (HT Jim Savage)

My impression – which I would be happy to have corrected – is that later research finds a smaller, but still significant impact of genetics on happiness.

That result brought to mind something John Stuart Mill wrote in his autobiography: “I am now convinced that no great improvements in the lot of mankind are possible until a great change takes place in the fundamental constitution of their modes of thought.” I imagine Mill had education and cultural shifts in mind, and the consistent difference in reported subjective wellbeing between South America and Eastern Europe show culture makes a difference. But if genetics plays such a large and limiting role, the only way to drastically alter our ‘modes of thought’ will require that we learn how to tinker with our minds directly.

GD Star Rating
loading...
Tagged as: , ,