Laura Carter

SoftBank is a Japanese investment conglomerate most famous—or notorious—in the English-speaking world for their string of investments in startups you've never heard of* (and companies like WeWork, which you have heard of for being wildly overvalued). A presentation by their founder and CEO, Masayoshi Son, for the 2026 Shareholder AGM** has been doing the rounds this week for being, frankly, completely deranged about AI.*** Using a metaphorical goose.

As best I can tell, the presentation makes the argument that the value of the company can be expressed in golden eggs. Over 16 years, SoftBank has gone from a valuation of 3 billion yen to 74 billion yen. Son's argument is that this value increase doesn't come from the eggs (slide 48 states that 'eggs do not lay eggs') but from the goose that laid them. And according to slide 55, 'ASI'—almost certainly 'artificial superintelligence,' a term that AI superfans use for hypothetical systems that are orders of magnitude smarter than humans—will be a goose that produces even more golden eggs.

Obviously, this is nonsense. ASI isn't real. And the golden egg metaphor obviously doesn't work: goose eggs don't lay more eggs, but rich people's money absolutely does turn into more money, through interest, dividends, and returns from selling stocks and so forth.

But this slide is inadvertently a pretty good encapsulation of how a lot of people—and unfortunately, many people in decision-making roles—think about 'artificial intelligence' in general. That's not how geese work! A goose isn't a factory with feathers, containing an assembly line producing eggs (golden or otherwise). It's a bird: it will lay only when it's old enough, when it's properly fed and cared for, and at the right time of year.**** Laying eggs isn't risk free: geese, like other birds, can get egg-bound if the egg gets stuck in the cloaca. And eventually the goose will get too old to lay.

It's also not how AI works: at least, not the AI we have today. Large language models aren't intelligent: they are enormous computer programs that use data and machine learning to model what human language looks like and produce samples of it based on the model. This is impressive! But it's not intelligence. There's a lot of money invested in making the argument that it is, and that soon LLMs will be as smart as humans, and soon after that exponentially smarter than us (that's the 'artificial superintelligence'). But LLMs aren't human, and they don't learn like humans, and they don't interact with others the way humans do, and there's no real evidence that they will any time soon (or indeed ever).

Factories also require maintenance and supplies, of course. But they're very different from those required by geese! Treating a goose like a factory ignores all the reasons why geese lay eggs and why factories produce products: just as treating a large language model like an intelligent human ignores all the reasons why LLMs produce generated text and humans produce...language, communication, creativity, relationships, and all the other signs of intelligence you can think of. Trying to get more eggs out of a goose with more electricity, better machinery and increasingly automated processes is not going to work. Nor is trying to get 'more intelligence' out of an LLM with more electricity, data and chips. However much tech companies and investment conglomerates would like to think they are the same thing, geese aren't factories, and humans aren't LLMs.

The original fable, of course, ends with the owners, driven by greed, killing the goose that laid the golden eggs: only to find that it's just the same as all their other geese on the inside. Treating LLM-generated text as intelligence, and therefore extrapolating a misunderstanding of where it comes from, what it can do, and what its limits are, is likely to end badly. Masayoshi Son and SoftBank might lose billions of yen, but they will still have plenty. The rest of us might not be so lucky.

* Someone I know worked at a startup which received SoftBank money. They told me the funding went “straight up the founder’s nose.”

** I have linked there to an archive.org copy of the presentation. There is currently a version of the presentation on the SoftBank website, though I haven't been able to find a direct link to it from the page about the 2026 AGM. The AGM was conducted in Japanese, and the slides are in English.

*** Even compared to some of the nonsense doing the rounds currently, this is ludicrous. Which is saying something.

**** Unlike chickens, which do lay all year round. And also unlike chickens, there aren't many goose egg factory farms (geese are factory farmed, but for their meat and for their livers in foie gras).

This year, I’m working through Nathalie Tasler’s prompts. I'll keep updating this post as the month goes on.

Day 9

Cheating a little here, with from the Scriberia illustrations commissioned during a Turing Way book dash, back when I was in the very early stage of thinking about this idea.

Day 8

Mastodon: I'm writing a book about how different people are—and are not—represented in different data sets. In particular, it's about how the ways that different actors collect data exclude or misrepresent people who are in some way out of the mainstream: because of their #gender, their way of life, their identity/ies.

LinkedIn: My current book project focuses on exclusion and misrepresentation in data sets. When we collect data about people, the information that we choose to collect is a choice: and not everyone fits neatly into predefined categories.

Day 7

I have a tendency to ramble in my writing, and to use a lot of run-on sentences. I'd like to get better at writing more concisely—and at editing my own writing.

Day 6

I don't take as much care of my readers as I should: at least, not in academic writing. Sometimes this is because I'm racing to meet a deadline and 'done' is better than 'perfect,' sometimes it's because I'm so fed up of a paper that I just want shot of it. And sometimes it's because I feel pressure to 'perform' academic inscrutability. I'm trying to move away from that—my book project idea is for a general reader—but it does creep in, especially when I've been reading a lot of academic literature that tends towards the pompous or the jargon-y.

Day 5

A possible abstract for a paper or conference submission:

Missing drafts Machine learning—including generative AI—relies on training data, whether it is in the form of structured data, unstructured text, or images. But this data didn't spring into existence fully-formed as a .html, .txt or .csv file. This paper will make visible the information and thought that went into creating the training data that is used to build machine learning, large language and image generation models. From court case submissions that are used in crafting the final judgement produced by a judge, to tried-and-failed experiments which never produce academic papers, to the previous drafts of novels which exist only on the writer's computer: this paper will show how this hidden information is vital to understanding how training data was created, and therefore to understanding the performance of machine learning models.

Day 4

To write, I need quiet, a door that closes, to be left alone. I need to spend time settling in to the work, to pick up where I left off before, and to quiet my brain from the day-to-day.

This isn't always feasible! But it's easier to come by in the afternoons.

Day 3

Weakness: sometimes it is useful to spend time looking for a reference! Future Laura isn't always grateful when Past Laura leaves her with all the work to do. Sometimes looking for the reference helps me realise that the point I was trying to make isn't actually that useful or relevant.

Strength: when I am writing well, I am able to put a lot of words down—and it's always easier to edit existing text than come up with it in the first place.

Day 2

When I am writing at my best, I am like a well-oiled machine: the words flow easily from my brain through my fingers and on to the screen. I don't get bogged down in looking for the perfect reference or exploring a footnote – I simply note where these are needed and keep going.

Day 1

I'm hoping to use these prompts to help me think through my current book project!

The public doesn’t want AI ‘mainlined into the veins’ of the UK: at least not when it comes to the public sector.

How do we know? Well, my former colleagues at the Ada Lovelace Institute have been asking them. Over the last six years, Ada has done a lot of research into what the public want from data and AI. Earlier this year, they asked me to help pull together findings from their work with 16,000 people in four nationwide attitudinal surveys and 400 people in deeper qualitative studies, to help answer the question of how the general public wants the UK public sector to use AI.

Here's what we found:

1. There is no single view of AI: public perceptions are nuanced and context-dependent. People assess tools in context and can identify the benefits, opportunities, risks and harms of individual AI use cases.

2. Experiences and demographics shape people’s expectations. People’s understanding, trust and comfort with AI are affected by their personal characteristics, as well as their direct and indirect experiences of technology and the institutions using it. These experiences can exacerbate concerns about AI’s impact on existing inequalities.

3. Private profits, public doubts: there are concerns about the power and profits of technology companies in public services. This intersects with concerns around transparency, regulatory powers and access to data.

4. Support is conditional: the public want evidence, explainability and involvement. The public value explainability above accuracy when it comes to AI tools, and they expect clear evidence on efficacy and impacts to justify use. There is a desire for those affected to have meaningful involvement in shaping decisions about public sector AI.

5. Strong governance is a prerequisite for trust. The public increasingly ask for stronger governance of AI, along with clear appeals and redress processes if something goes wrong. The public are not convinced existing regulations are adequate to ensure that public sector AI works for everyone, supports public good and prioritises people over profit.

6. Social impact matters: the public oppose uses of AI that could create a ‘two-tier society’. Discrimination and bias are important concerns, especially in essential services, and people expect AI in the public sector to accommodate pluralism and diversity.

The public is more than capable of understanding different use cases, implementations and outcomes, and they care if, where and how public sector AI is used. Public legitimacy matters: if the UK government wants to roll out AI across the public sector, it is important that it’s done properly.

You can read the whole report here.

I’ve had a lot of thoughts in the wake of the UK Supreme Court’s decision a couple of weeks ago to define women and sex in terms of biology. Others have said better than me that the judgement is transphobic, it will hurt (indeed, has already hurt) people, and that it is a misreading of everything that the Equality Act stands for. Many people—trans and cis—have been protesting the decision, and a judge plans to take the UK to the European Court of Human Rights.

As a human rights expert who started working on legal gender recognition back around 2012, I can say that it is a bad judgement, and I hope that the ECtHR recognises that: but the process will take years. In the meantime, protections against discrimination on the basis of 'transgender status' still exist in the law, but decision-makers are using this ruling to exclude trans people. Toilets—predictably, for anyone who has followed this issue and been exposed to the inherent creepiness of people thinking about the genitalia of the person in the next cubicle—are one area of exclusion. But another area is sport. Excluding trans women, apparently, is necessary for the safety of women and girls.

I was a sporty girl. I wasn't exceptional at any particular sport, but a height advantage (I was taller than most of my classmates from about the age of nine), generally good spatial awareness, and a decent ability to follow instructions and pick up rules quickly made me a competent B-team player in most of the sports I tried: and I liked sports, so I tried most of them. That brought me into contact with a whole load of coaches.

Let's talk about three of them. I'm going to call them Pat, Paul, and Mr S.

Pat taught swimming at my secondary school.

Paul coached the rowing team at my secondary school.

Mr S. was another rowing coach, working for a club we sometimes competed against. Paul and Mr S. also coached elite junior teams in the summer, along with several other school coaches, Some of my teammates were selected for those teams (not me. I remained a B-team player my whole adolescence): some ended up in boats coached by Paul, some by other coaches.

Once a week, starting from year seven, Pat herded my class of girls (it was a girls' school) into the changing rooms to get changed into their swimsuits.

By the summer term, the rowing teams for the junior European and World championships had already been selected, but we were still racing as a school team to finish out the season. We'd sit in the minibus (rowing competitions involved a lot of sitting in minibuses between heats), eating carbohydrates, pointing out who was on which team, and gossiping about their chances.

At one of these competitions, I was sitting in the minibus with Paul and another girl, H. Paul was in the front, doing something with paperwork, and not really listening to us talk. H pointed out a man walking past. That's Mr. S, she said. He's coaching some of the elite teams this summer. Not ones that any of us are in. Which is good, she said, because I've heard he's kind of creepy with the girls he coaches.

Once we were all in the swimming pool changing rooms, Pat...stayed. Sitting on a bench at the side of the classroom, watching as we changed. Watching us take off our uniforms and put on swimsuits. It's been more than 20 years, and I could still draw you the layout of the changing room, and show you exactly where Pat sat and watched.

Paul was listening, it turns out. I can also remember exactly what he said to H and me in the minibus that afternoon. Mr S. isn't coaching any of you, Paul said, because I put my foot down and said I didn't want him coaching any of my girls. Because he's creepy? H asked. Yes, Paul said, and I don't want him around any of you.

This was in the late 90s and early 00s in the UK. There are better laws and policies safeguarding children now. I think—I hope—it would be harder for people like Pat and Mr S. to slip through the net. To abuse their positions as adults working in children's sports, to invade their privacy and make them feel like objects.

Mr S. was a cis man, perhaps in his 40s then. I never actually spoke to him. Thanks to Paul, I never had to.

Pat was a cis woman in her 50s. Thanks to her, I spent years avoiding swimming as an activity, as well as avoiding changing rooms. I'm still not a strong swimmer.

Paul was a cis man in his early 20s: at most seven or eight years older than the girls he was coaching. He had maybe two or three years of coaching experience at that point, and was probably 20 years younger than Mr S. I don't know how much career capital he had at the time, or how much he had to argue to keep us away from Mr S. But I know that he did it.

I don't know where Pat is now, nor Mr. S. I don't want to find out. Maybe they're dead. I hope they're no longer anywhere near adolescents.

I do know where Paul is. I doubt he remembers that day in the minibus at all. If he remembers me at all, it's as one of a gaggle of (probably quite annoying) teenagers he used to coach more than 20 years ago. He was a great coach—he led our B-team of teenagers to a large number of victories, including over teams of fully-grown adults—but more than that, he made it clear that predatory, pervy behaviour from coaches wasn't acceptable, and that he was willing to stick his neck out to protect us from it. And so I do remember him, and occasionally I look him up to see where he is now. His name actually is Paul. He's apparently still a great coach, too: he's an Olympic Head Coach for British Rowing.

Trans people aren't a threat to girls in sport. Cis people like Pat and Mr S. are a threat. No-one—cis or trans—should have to put up with ogling, or with their privacy being invaded, in order to play sports. Robust, effective measures to keep predators out of sports are what is needed to keep women and girls safe. Cis and trans women and girls just want to be able to play.

There is scientific consensus that AI can exacerbate bias and discrimination in society: even when its developers have the best of intentions. I'm proud to be one of the more than 140 academic researchers to sign a statement affirming this consensus and stating in no uncertain terms the need for appropriate mitigations to the harms that this bias and discrimination can cause.

We, the undersigned researchers, affirm the scientific consensus that artificial intelligence (AI) can exacerbate bias and discrimination in society, and that governments need to enact appropriate guardrails and governance in order to identify and mitigate these harms.

Over the past decade, thousands of scientific studies have shown how biased AI systems can violate civil and human rights, even if their users and creators are well-intentioned. When AI systems perpetuate discrimination, their errors make our societies less just and fair. Researchers have observed this same pattern across many fields, including computer science, the social sciences, law, and the humanities. Yet while scientists agree on the common problem of bias in AI, the solutions to this problem are an area of ongoing research, innovation, and policy.

These facts have been a basis for bipartisan and global policymaking for nearly a decade. We urge policymakers to continue to develop public policy that is rooted in and builds on this scientific consensus, rather than discarding the bipartisan and global progress made thus far.

The full list of signatories is here: https://www.aibiasconsensus.org/

I’ve been training to become a trainer with the Digital Rights Academy. As part of this, I prepared and delivered a one-off training session: an introduction to the reference management software Zotero.

Zotero is open source, free,* and made my life much easier when reading for and writing up my PhD (which ended up with 1,000+ citations…). But it can also be a bit intimidating for new researchers, especially for students in humanities subjects who might never have used any kind of research software in the past. So I designed this training to be as accessible as possible, and to start from the beginning: how to add information about your references (metadata), how to use Zotero’s plugins to add research papers directly from journal webpages, and how to use it in Word to generate citations.

The slides from the training are available here if you’d like to see the information that we covered. And if you’d like to see the DRA run this again, please do get in touch!

* If you use it a lot, you may end up wanting to pay for increased storage—I did—but the free version gives you enough storage to play around with and find out if you like using it. I recommend not upgrading until you need to.

The Feminist Review Trust was founded in 2001. The Feminist Review, an academic journal, was making a profit: the Trust was a way for a small group of socialist feminists to redistribute those profits into feminist projects worldwide. Over two decades, the Trust distributed over half a million pounds to nearly 200 projects, across five continents.

I became a Trustee in 2017. The Trust never had any paid staff: it was the job of the Trustees to review the hundreds of grant applications we received each round, and then, three times a year, to meet to deliberate over which ones we could fund. Each funding round, I sat with my colleagues—feminists from different generations, backgrounds and experiences to me—as we deliberated over which projects to fund.

It wasn’t ever an easy process. We never had enough money to fund everything that we wanted to. We prioritised projects that we thought would struggle to find support elsewhere, and feminists in countries with less infrastructural support. We didn’t look for scalability, or even always ongoing work: we wanted to fund transformative change, and that didn’t always mean big projects. We trusted that our grantees knew what was needed in different contexts, and tried not to burden them with too many reporting requirements.

We didn’t always get it right. Occasionally, grantees would vanish after receiving the funds (but not nearly as often as we might have feared). Sometimes they wouldn’t be able to deliver on the project that they proposed, but almost always they’d come back with justifications for the delay, or reasons why they wanted to reallocate our funding. We almost always said yes.

When the Trust closed its doors, we commissioned two reports. One looking backwards: a history of the Trust. And one looking forwards: an evaluation of what we did and what we learned from it, in the hope that it might be useful to other (and future) feminist funding. There is more support for feminist organising now, and different models of funding: but still not nearly enough.

I'm proud to have been a Trustee. I learned an enormous amount from my fellow Trustees, and from the feminists that we funded over the years. I'll miss our meetings and my colleagues, and I'll miss finding out about the amazing work that feminists around the world were able to do with the (really, quite tiny amounts of) money that we were able to send to them.

Our papers are archived at LSE, and copies of both reports are at the Copyright Libraries. When we finished the winding down of the Trust, we had a small amount of money left over. In our final political feminist act, we donated it to Medical Aid for Palestinians.

Over on the OLS* slack, the community coordinator came across—and reposted—a blog post that I wrote a couple of years ago: The Human Rights Case for Open Science.**

It prompted a question from another OLS participant, who had had unsatisfying experiences as a citizen scientist:***

If participants in citizen science contribute their labour to support the project, but it only benefits the project organisers, is this a human rights issue?

I've been working in the human rights field for more than a decade now, so I tend to think everything is a human rights issue, in the sense that you can use a human rights lens to look at the question.

The short answer is yes: but the human rights framework might not be the best way to fix the problem.

The bit of human rights law to start with is probably Articles 6-8 of the International Covenant on Economic, Social and Cultural Rights, which cover the rights to work, to just and favourable conditions of work, and to form trades unions. It's a bit tricky to apply this to citizen science projects though, where participation ought to be voluntary – not just in the sense of unpaid (though this is often the case!), but also that people can stop being involved at any point.

Citizen scientists who don't like a project should just be able to walk away. It's on the project organiser to set up a project that people want to be involved in, whether because they think the issue is important, they want to learn about doing science, or they want to be part of a community, or any other reason.

Problems arise when project organisers don't deliver on the benefits that they promised, or when citizen scientists can't walk away. If this rises to the level of coercion—for example, withholding services unless someone agrees to do data collection work—then in some circumstances this might rise to the level of a human rights abuse. I’m not aware of any community science projects where this has happened! This kind of abuse is much more common in other industries, for example agriculture or construction: you can read more about this in the work of organisations like Anti-Slavery International.

But more often it's going to look like a project organiser taking credit for the work of others, or reaping the benefits of a successful project (in prestige, citations, or grants) while the citizen scientists don't benefit. Or sometimes, it might look like a bullying project organiser making someone feel guilty if they want to step away. Those kinds of situations don't feel nice for the participants, but they are unlikely to rise to the level of human rights abuses.

So in those kinds of situations, the best recourse probably isn't going to be the human rights system (fond as I am of it). If you want to stay involved with a project,**** your best option for improving it is probably going to be organising.

Unionisation is protected by Article 8 of the ICESCR, but there are lots of other ways to work for change. There's a chapter in the Turing Way***** on activism for researchers: it's written for professional data scientists, but almost all of it is applicable to anyone who wants to change how they do science for the better.

* I've been a participant, and a couple of times a mentor, for OLS's Open Science training, which I highly recommend for anyone interested in science and community building.

** The blog was originally written for Open Heroines who are also worth checking out if you're a women or non-binary person interested in open data.

*** I initially wrote an answer on the OLS slack. But slack messages are private and transient: this is an edited, expanded, and public version!

**** But you don't have to! Volunteer Amnesty Day exists for a reason.

***** The Turing Way is a guide to collaborative, reproducible and ethical research: there's a lot in here about how to do collaborative science well! I was a contributor to the Guide for Ethical Research—in fact, that was my first OLS project!—and wrote some of the Activism content. If you're interested in getting involved, I recommend starting with the contribution guidelines.