
ALT Checks on the industry are further compromised by the close ties between tech companies and ostensibly independent academic institutions. Researchers from corporations and academia publish papers together and rub elbows at the same conferences, with some researchers even holding concurrent positions at tech companies and universities. This blurs the boundary between academic and corporate research and obscures the incentives underwriting such work. It also means that the two groups look awfully similar—AI research in academia suffers from the same pernicious racial and gender homogeneity issues as its corporate counterparts. Moreover, the top computer science departments accept copious amounts of Big Tech research funding. We have only to look to Big Tobacco and Big Oil for troubling templates that expose just how much influence over the public understanding of complex scientific issues large companies can exert when knowledge creation is left in their hands.

ALT Carceral technologies are racist because the institutions that develop and use them are intended to manage populations in a country that has a white supremacist inheritance. These technologies are not incidentally racist. They are racist because they're doing the work of policing—which, in this country, is a racist job. There has been a lot of work devoted to proving that particular algorithms are racially biased. And that's well and good. But there was no question that these algorithms were ever not going to be racist. What would a not-racist predictive policing program look like? You would have to reimagine prediction. You would have to reimagine policing. You would have to reimagine the history of computation. You would have to reimagine the racial configuration of neighborhoods. You would have to reimagine a lot of things in order to arrive at even the slightest possibility of a not-racist predictive policing system, or a not-racist facial recognition system. So yes, they're raci

ALT Feeling completely stuck One very common challenge is feeling completely stuck, either on a specific project or with regards to the research process on the whole. If the challenge is a specific project, where you’ve pushed hard and it’s still not quite working, then some strategies that might help are Making a write up: Collect any partial experimental results, mathematical insights, jotted notes on motivation, etc and take time to put together a write up. This can help with providing a picture of where things stand and where the important gaps are. Pivot: if there’s a specific part of the project that’s not working, is there the possibility to reframe the question (possibly taking inspiration from related work) to make it more tractable? Forming connections: are there links between what the current project focuses on and other areas of study? Can that connection be explored in this project? This can both help progress and in making the project relevant to a broader community. Feedba

ALT In a talk entitled “How to recognize AI snake oil”, Professor Arvind Narayanan created a useful taxonomy of three types of tasks AI is commonly being applied to right now: Perception: facial recognition, reverse image search, speech to text, medical diagnosis from x-rays or CT scans Automating judgement: spam detection, automated essay grading, hate speech detection, content recommendation Predicting social outcomes: predicting job success, predicting criminal recidivism, predicting at-risk kids The above 3 categories are not comprehensive of all uses of AI, and there are certainly innovations that span across them. However, this taxonomy is a useful heuristic for considering differences in accuracy and differences in the nature of the risks we face. For perception tasks, some of the biggest ethical concerns are related to how accurate AI can be (e.g. for the state to accurately surveil protesters has chilling implications for our civil rights), but in contrast, for predicting social


















