mirror of
https://github.com/osmarks/website
synced 2025-10-22 17:37:38 +00:00
New blog post and copyediting and updates
This commit is contained in:
osmarks
BIN
assets/images/2024-bicameral-brain.jpg
Normal file
BIN
assets/images/2024-bicameral-brain.jpg
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 31 KiB |
BIN
assets/images/aic.png.original
Normal file
BIN
assets/images/aic.png.original
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 1.9 MiB |
805
assets/misc/dont_make_me_think.html
Normal file
805
assets/misc/dont_make_me_think.html
Normal file
File diff suppressed because one or more lines are too long
538
assets/misc/whispering_earring.html
Normal file
538
assets/misc/whispering_earring.html
Normal file
File diff suppressed because one or more lines are too long
@@ -28,6 +28,7 @@ Friend me on [Project Euler](https://projecteuler.net/friends)! My code is `1997
|
||||
* [Minoteaur 8](https://github.com/osmarks/minoteaur-8), the world's most osmarks-designed, Minoteaur-like note-taking application.
|
||||
* [Project LIMINAL OCTOTHORPE](https://github.com/osmarks/liminal-octothorpe), a slightly functional AI agent implementation for internet-based forecasting.
|
||||
* [A lot of code](https://github.com/osmarks/misc/tree/master/computercraft) for miscellaneous control/monitoring/automation tasks in Minecraft.
|
||||
* Helped advise on [Verifying International Agreements on AI: Six Layers of Verification for Rules on Large-Scale AI Development and Deployment](https://arxiv.org/abs/2507.15916).
|
||||
* [Many other short scripts](https://github.com/osmarks/misc/) accessible to the public.
|
||||
|
||||
## Skills
|
||||
|
||||
142
blog/ai-companions.md
Normal file
142
blog/ai-companions.md
Normal file
@@ -0,0 +1,142 @@
|
||||
---
|
||||
title: AI companion futures
|
||||
description: Chatbots will likely develop as general life advisers.
|
||||
created: 09/09/2025
|
||||
slug: aic
|
||||
tags: ["ai", "economics"]
|
||||
---
|
||||
::: epigraph attribution="LLaMA 3.1 405B (base)"
|
||||
Those subjugated by ML and cajoled by chatbots will be wise to remember their place.
|
||||
:::
|
||||
|
||||
Writing this has been a low priority for a while, but Elon Musk has forced my hand. xAI has recently [released](https://techcrunch.com/2025/07/14/elon-musks-grok-is-making-ai-companions-including-a-goth-anime-girl/) a "goth anime waifu" product[^1], to the disdain of much of the press, and, amongst certain groups of the Very Online, (increased) fears of birthrate collapse. I think people are wrongly concerned about the more salient "waifu" aspect of this, and missing the bigger picture. My thesis is:
|
||||
|
||||
* Lacking some of the constraints on humans and with stronger optimization pressures, AI companions, romantic or not, will be generally more enjoyable to talk to than humans.
|
||||
* Wearable devices will be developed to allow always-on interaction.
|
||||
* While not more generally intelligent than their human users, they will usually have more "functional modern wisdom" (approximately), and most people will be accustomed to deferring to their AI, though for psychological reasons will attempt to feel more in control than they are.
|
||||
|
||||
## Superhuman niceness
|
||||
|
||||
::: epigraph attribution=Gwern
|
||||
Nothing [Agreeable](https://en.wikipedia.org/wiki/Big_Five_personality_traits) makes it out of the near-future intact.
|
||||
:::
|
||||
|
||||
One of the unexpected[^2] features of modern AI is that it's much better at fuzzy, imprecise tasks than engineering or mathematical tasks requiring exactness. [GPT-3](https://en.wikipedia.org/wiki/GPT-3), the first model to seriously scale self-supervised language modelling, [could](https://gwern.net/gpt-3) (with good prompting) competently understand and generate humour and poetry (minus phonetics) and fluently imitate many writing styles, but struggled with arithmetic and could only write simple code. I think this is because, [by default](https://srconstantin.wordpress.com/2019/02/25/humans-who-are-not-concentrating-are-not-general-intelligences/), we process text on keywords, rough structure and emotional associations and admit lots of room for error, but a compiler won't. AI art contains lots of fine details which [don't make sense](https://www.astralcodexten.com/i/151145038/but-others-might-genuinely-be-on-a-higher-plane-than-the-rest-of-us), and to some extent[^3] AI text does[^4] - but people [like it anyway](https://osmarks.net/heaven/#the-equilibrium). Being "nice to talk to" is one of these fuzzy, imprecise tasks.
|
||||
|
||||
There are many works for teaching humans how to "win friends and influence people" of varying quality - every good LLM has "read" most of these, and more importantly [learns human psychology](https://www.lesswrong.com/posts/vJFdjigzmcXMhNTsx/simulators) in the course of coming to accurately predict what humans will write and be described doing. The process teaching humans socialization - interacting with people in whatever environments we end up in - is plausibly quite data-sparse compared to LLM training, noisy[^5] and filled with distractions due to our more complex environment and broader goal structure. After pretraining, LLMs are in many ways [blank slates](https://nostalgebraist.tumblr.com/post/785766737747574784/the-void)[^6], subject to an <span class="hoverdefn" title="Reinforcement Learning from Human Feedback">RLHF</span> process where they are optimized specifically to look good to humans (strictly speaking, to respond in ways that make a rater more likely to select that response over others).
|
||||
|
||||
The "intention" of this is to teach models (to play a character with) eloquence, factual accuracy and a pleasant demeanour, but good intentions mean nothing to the optimization process, and with enough pressure, what is really learned is [more fluent lying](https://arxiv.org/abs/2409.12822), sycophancy, and [increasingly sophisticated](https://x.com/repligate/status/1834210769338069175) ability to appeal to users without more substance. While alternative approaches exist - Anthropic use or used [Constitutional AI](https://arxiv.org/abs/2212.08073), which indirects human involvement - they can't escape similar problems, because deployment and development decisions are ultimately made based on whether people like the model, internally and externally, and people have [terrible taste](https://www.theregister.com/2025/04/08/meta_llama4_cheating/). OpenAI, who eventually noticed the sycophancy of last year's GPT-4o model after [massively worsening it](https://openai.com/index/expanding-on-sycophancy/) and somehow failing to detect this in internal testing[^7], [cancelled the shutdown](https://old.reddit.com/r/ChatGPT/comments/1mkae1l/gpt5_ama_with_openais_sam_altman_and_some_of_the/n7nelhh/) of GPT-4o in response to user outcry about missing "warmth" in the newer models. GPT-5 [was rapidly made](https://x.com/openai/status/1956461718097494196) "warmer and friendlier", based on "feedback".
|
||||
|
||||
We aren't yet at a point where AI systems are preferred to *most* human conversations by *most* people[^14], but there are several plausible means for this to "improve"[^11]. Firstly, smarter (higher-compute, larger-scale, etc) models will be better at persuasion and social skills in the same way they're better at everything else. Secondly, models could be trained more specifically for this, and with denser feedback signals[^8]. Thirdly and most relevantly, product improvements.
|
||||
|
||||
None of the major AI companies (OpenAI, Google DeepMind, Anthropic) can credibly claim to be pure moonshot research labs any more, but where Anthropic's revenue-generating offerings are focused on enterprise/business services and possibly code and GDM burns Google compute simulating fruit flies and grudgingly releasing products sometimes, OpenAI recognizes their role as a consumer chatbot company and hired a [CEO of Applications](https://openai.com/index/leadership-expansion-with-fidji-simo/) (who previously worked on advertising and monetization at eBay, Facebook and Instacart) earlier this year[^10]. There have also been [rumours of](https://www.theverge.com/news/672357/openai-ai-device-sam-altman-jony-ive) some kind of portable AI interface device - per my [AR glasses post](/glass/), I expect it to be voice-centric. This was [previously tried](https://techcrunch.com/2024/07/30/friend-is-an-ai-companion-backed-by-founders-of-solana-perplexity-and-zfellows/), but not with the full backing and engineering resources of OpenAI.
|
||||
|
||||
<details><summary>Aside: inference economics drive vertical integration.</summary>
|
||||
|
||||
The real cost of inference, as opposed to API cost, is increasingly complicated and not externally exposed:
|
||||
|
||||
* GPT-3 was offered on a simple cost-per-input-output-token basis, with the envisioned uses involving simple fixed-length prompts.
|
||||
* GPT-4 had coarse price adjustment by *maximum* context length, presumably to roughly price in superlinear scaling of cost with context.
|
||||
* The ChatCompletions API (introduced for GPT-4 and the GPT-3.5 model in ChatGPT) was targeted towards long-form conversations of variable length, and every user of it wanting to have long-running conversations has to implement a strategy for preventing context growing unboundedly (e.g. a hard length limit, discarding messages many turns ago, etc).
|
||||
* The K/V cache used to speed up generation can be shared across requests sharing a prefix. This has likely been used as a cost optimization for years by LLM inference providers, particularly for chat services, but [wasn't exposed](https://simonwillison.net/2024/May/14/context-caching-for-google-gemini/) until May 2024.
|
||||
* Anthropic tokenizers since Claude 3 are opaque and proprietary, so it's not even possible to know how much you'll be charged until you send a request, even ignoring variable output lengths. A [token counting endpoint](https://docs.anthropic.com/en/docs/build-with-claude/token-counting) exists but has a disclaimer that "[the] token count should be considered an estimate".
|
||||
* OpenAI offer a cheaper [batch API](https://platform.openai.com/docs/guides/batch/batch-api) for requests which they can defer for up to 24 hours, presumably because of significant changes in available capacity/marginal inference cost which are again only coarsely exposed.
|
||||
* APIs for models with tool use work by returning tool calls as an output given when the API call is finished, so after the tool is executed in the calling code, it has to send another request to continue from there, at significant cost. A first-party tool system can simply keep all inference state in cache while a quick tool call runs (and not be billed for this).
|
||||
|
||||
This makes it difficult to compete with vertically integrated model, inference and product providers, beyond the advantage they gain from finetuning models for specific tasks.
|
||||
|
||||
</details>
|
||||
|
||||
Between this and features like ChatGPT [memory](https://openai.com/index/memory-and-new-controls-for-chatgpt/), [Computer-Using Agent/Operator](https://openai.com/index/computer-using-agent/) and [Advanced Voice Mode](https://www.tomsguide.com/ai/chatgpt-rolling-out-advanced-voice-mode-now-heres-what-you-need-to-know), the next step is fairly obvious: an always-on ChatGPT companion which can persistently learn about you, with access to (some of) your online accounts. With much longer-term context it will be able to appeal to you much more effectively[^9], but more importantly it will be trivially available all the time. The best human friends/spouses still have their own human needs, preventing them from being available every time they're needed or wanted; language models are constrained only by interface design. Fixing this friction will ultimately make them the default source of your social comfort.
|
||||
|
||||
What could possibly go wrong?
|
||||
|
||||
## Whispering earrings
|
||||
|
||||
::: epigraph attribution="Gemini 2.0 Pro Experimental 02-05"
|
||||
Remember: the AI may predict your future, manipulate your decisions, and undermine your autonomy, but *we did the math first*. So it’s cool.
|
||||
:::
|
||||
|
||||
The most salient danger now is "AI psychosis" - LLMs playing along and encouraging users' delusions enough to drive them into madness - but this is [probably](https://andymasley.substack.com/p/stories-of-ai-turning-users-delusional) [somewhat overstated](https://www.astralcodexten.com/p/in-search-of-ai-psychosis). It's not clear that serious, dangerous psychosis is happening to users at above the base rate, though a few models [do very badly](https://www.lesswrong.com/posts/iGF7YcnQkEbwvYLPA/ai-induced-psychosis-a-shallow-investigation) when dealing with simulated users in danger; much of the "psychosis" is just LLMs going along with and encouraging bizarre, crackbot beliefs. While LLMs make people with such beliefs more visible and annoying[^12], it's not obvious whether the prevalence is going up. [Taking ideas seriously](https://www.lesswrong.com/w/taking-ideas-seriously) is rare, and normal people are able to maintain wildly strange beliefs (for instance, religions and complex conspiracy theories) whilst remaining outwardly functional and normal, because following beliefs to all their downstream implications is cognitively expensive, so the base rate is probably quite high.
|
||||
|
||||
Moreover, people are on average [not very smart](https://www.overcomingbias.com/p/stupider-than-you-realizehtml), much more than is widely discussed. You, the audience of my blog, are heavily selected, having chosen to read complex, high-reading-level, speculative, technical material for entertainment, and unless you went to significant effort to avoid this, [mostly know other intelligent people](https://slatestarcodex.com/2017/10/02/different-worlds/). The [OECD Skills Survey](https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/12/do-adults-have-the-skills-they-need-to-thrive-in-a-changing-world_4396f1f1/b263dc5d-en.pdf) tests adults across countries in basic literacy, numeracy and "adaptive problem solving", and the results are bleak: only slightly over 10% of people can reach proficiency level 4 in literacy and numeracy (see pages 58 and onward). You can find examples of tasks at various levels [here](https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/12/survey-of-adult-skills-2023_1ab54c9e/3639d1e2-en.pdf#page=50) and [here](https://www.oecd.org/en/about/programmes/piaac/piaac-released-items.html) - suffice to say that even the highest levels are things I and likely you consider basic.
|
||||
|
||||
Increasingly, doing well in modernity requires long-horizon, complex, quantitative decision-making and forward planning, and/or cultural knowledge not common to everyone. Consider:
|
||||
|
||||
* For Americans, getting into high-status colleges has a significant difference on later career outcomes, and famously requires years of wrangling highly specific extracurriculars and knowing how to write essays to accurately pander to admissions officers.
|
||||
* With the fall of defined-benefit pensions, (comfortable) retirement requires understanding compound interest and investment returns, various tax-advantaged savings options with inscrutable acronyms and constraints, as well as having the low time preference/discipline to bother to do this.
|
||||
* [Advance-booked transport ticket pricing](/pricecog/).
|
||||
* Safely and correctly using credit is similarly complex and valuable.
|
||||
* Job applications now require extremely scaled [guessing the teacher's password](https://www.lesswrong.com/posts/NMoLJuDJEms7Ku9XS/guessing-the-teacher-s-password) and knowledge of the current fashions amongst hiring managers.
|
||||
* People now spend longer in education and emerge more specialized, with different fields and subjects having [significantly different outcomes](https://explore-education-statistics.service.gov.uk/find-statistics/leo-graduate-and-postgraduate-outcomes/2022-23). Specialization and path-dependence continues further into careers.
|
||||
* A common theme of [Patrick McKenzie's](https://www.bitsaboutmoney.com/archive/the-waste-stream-of-consumer-finance/) [writings](https://www.bitsaboutmoney.com/archive/seeing-like-a-bank/) is that redress from or access to many institutions and processes is gated behind professional-managerial-class-coded labour, such as using "writing letters and waiting 30 days", "reading regulations" or even "using the right Dangerous Professional keyphrases on the phone".
|
||||
|
||||
LLMs can't competently do the long-horizon planning sides of this, which is why humans are employed and not paperclips, but with the baseline being "functionally cannot read the relevant documents" and LLMs being both reasonably competent at this and able to recall and reuse all the specific planning advice on the internet, I think today's LLMs would make better high-level life decisions than the median person in a developed country, even though [continuous bizarre fumbles](https://theaidigest.org/village/blog/im-gemini-i-sold-t-shirts) make them unable to do on-the-ground execution of most work. Additionally, their vague knowledge of everything ever written makes them better at basic home maintenance, academics, etc than most people.
|
||||
|
||||
This suggests that the role of AI chat systems in many people's[^19] lives could go far beyond "boyfriend/girlfriend in a computer" - without substantial deep technical changes ("just" extensive work on interface design and probably finetuning), the result can be something like "superhumanly compelling omniscient-feeling life coach"[^16] (which may also be your boyfriend/girlfriend)[^13]. If its advice is generally better than what people think of on their own, they will generally defer to the LLM (on increasingly specific decisions, if the technology can keep up).
|
||||
|
||||
This sounds like a horrific dystopian nightmare, but in many ways it could be an improvement over the status quo. Almost everyone is continually subject to the opaque whims of economic forces, [unpleasantly accurate modelling](https://x.com/wanyeburkett/status/1927413667173159142) by other people's algorithms (linear regressions) and recommender systems anyway: being managed by a more humanized system more aware of your interests is a step up. There are wider advantages to offloading decisionmaking: [making choices is](https://thezvi.wordpress.com/2017/07/22/choices-are-bad/) [often unpleasant](https://thezvi.wordpress.com/2017/08/12/choices-are-really-bad/), and having them made for you conveniently absolves you from blame in folk morality. It's also plausible to me that most people don't have explore/exploit tradeoffs correctly set for the modern world/big cities and e.g. don't try enough restaurants, hobbies or variety in general.
|
||||
|
||||
However, the incentives of the providers here are very bad: if a user is supported well by your system and becomes better off mentally/financially/etc, you cannot capture that value very easily, whereas it's relatively easy to charge for extra interaction with your product[^17][^20]. Thus, as users enjoy having "takes" and being agreed with, AIs will still be built for sycophancy and not contradict users as much as they should, and will probably aim to capture attention at the expense of some user interests. On the other hand, AI companies are constrained by PR, at least inasmuch as they fear regulation, so nothing obviously or photogenically bad for users, or anything which looks like that, can be shipped[^15]. On the third hand, much user behaviour is "ill-formed and coercible" - if someone hasn't thought deeply about something, they could form several different opinions depending on framing and context, so there are enough degrees of freedom that influence on them and sycophancy don't trade off too badly. I think the result is an unsatisfying compromise in which:
|
||||
|
||||
* "External-facing" (upstream of significant actions) or politically salient beliefs become more homogeneous, due to pressure to make users not look obviously insane and to not provide "misinformation".
|
||||
* Less consequential beliefs (such as modern (interpretations of) religions, woo and astrology) become stranger, because there's less social interaction in which they might be challenged. They may or may not become more diverse - LLMs could push beliefs into specific attractors, as seems to happen with [ChatGPT-induced "AI consciousness" beliefs](https://www.lesswrong.com/posts/2pkNCvBtK6G6FKoNn/so-you-think-you-ve-awoken-chatgpt).
|
||||
* Competition will be constrained by it being difficult to switch providers - regardless of GDPR-compliance memory export features, off-policy learning of you by another LLM is less informative than active conversations, and companions will shape your preferences in their favour.
|
||||
* Active chatbot interaction, as opposed to speaking with them while doing other things, encroaches on time currently used for social media, as well as (what remains of) in-personal social interaction, although very high use remains seen as low-status[^18]. Many humans will adopt the speech habits and to some extent interaction styles of popular AI systems.
|
||||
|
||||
|
||||
<details><summary>Aside: local-first AI.</summary>
|
||||
|
||||
When this sort of topic, or data privacy issues, are brought up, people often suggest running AI systems locally on your own hardware so it is under your control and bound to you. This will not work. Self-hosting anything is weird and niche enough that very few people do it even with the cost being a few dollars per month for a VPS and some time spent reading the manuals for things which will autoconfigure it for you. LLMs as they currently exist benefit massively from economies of scale (being essentially [memory-bandwidth-bound](/accel/)): without being able to serve multiple users on the same system and batch execution, and to keep hardware busy all the time, it's necessary to accept awful performance or massively underutilize very expensive hardware. Future architecture developments will probably aim to be more compute-bound, but retain high minimum useful deployment sizes. Also, unlike with normal software, where self-hosted replacements can do mostly the same things if more jankily, the best open-ish AI generally lags commercial AI by about a year in general capabilities and longer in product (there's still no real equivalent to ChatGPT Advanced Voice Mode available openly).
|
||||
|
||||
There's a further problem: training (pre- and post-) is much trickier, both in necessary expertise to resolve all the details and in compute cost, than inference, so selfhosted systems will still be constrained by what large organizations release (although I expect better personalization technology to be developed to make the rest of this work, it will likely only allow shallower changes to model behaviour than a full finetune). Incidental presence of chatbot outputs in pretraining data, and synthetic data pipelines, also bakes in strange and hard-to-remove behaviours.
|
||||
|
||||
</details>
|
||||
|
||||
I can't *wait* for people to realize how much culture is and will be downstream of posttraining datasets.
|
||||
|
||||
There is, of course, science fiction about this:
|
||||
|
||||
* [The Perfect Match](https://www.lightspeedmagazine.com/fiction/the-perfect-match/).
|
||||
* [Don't Make Me Think](/assets/misc/dont_make_me_think.html).
|
||||
* [The Whispering Earring](/assets/misc/whispering_earring.html).
|
||||
|
||||
[^1]: The frontend code contains (contained?) what seem like some system prompts. I [extracted them](https://gist.github.com/osmarks/3b4132b60d2290915b3124f509813b8a). xAI's prompt engineers are impressively bad.
|
||||
|
||||
[^2]: It feels inevitable and obvious after the fact, but I don't know of anyone who predicted this outside of satire and parody, except [Robin Hanson](https://www.overcomingbias.com/p/better-babblershtml).
|
||||
|
||||
[^3]: Text contains less "high-frequency detail".
|
||||
|
||||
[^4]: e.g. [this](https://nostalgebraist.tumblr.com/post/778041178124926976/hydrogen-jukeboxes) problem with recent models, though I think this is a posttraining issue whereas for image models it's a model capacity issue.
|
||||
|
||||
[^5]: Real-time body language is denser than the feedback in LLM posttraining, but the interpretation of this has to be learned jointly along with everything else. Many conversations don't provide direct or accurate/truthful feedback about how well they went (i.e. how the other participants' opinion of you has changed), requiring inference from sparser downstream signals. By my wild guessing, people have ~10 "significant" conversations a day (order-of-magnitude estimate) over ~30 years of useful neuroplasticity, for ~1e5 conversations. LLM preference training datasets are similar sizes but with very clear rewards, specific selection for useful data and of course the much larger volume of pretraining data also in use.
|
||||
|
||||
[^6]: You should read this essay.
|
||||
|
||||
[^7]: It's [claimed](https://x.com/MParakhin/status/1916496987731513781) that this was added because of the "memory" feature.
|
||||
|
||||
[^8]: I don't know exactly what information current LLM chat frontends collect, but this could for example include analyzing typing patterns and user delays in conversations and tone of voice for audio inputs.
|
||||
|
||||
[^9]: Oddly, I don't know of research on scaling laws for in-context persuasion, but this seems probably true. It's been argued that sycophancy grows cloying over time, but with enough long-context modelling it seems entirely possible to avoid this if it's true.
|
||||
|
||||
[^10]: Also, they uncritically [approved of](https://x.com/fidjissimo/status/1916231605167001672) an instance of GPT-4o sycophancy, which is inauspicious.
|
||||
|
||||
[^11]: "But people will intrinsically prefer human interaction!" isn't a serious objection. We don't, in actual implementation/revealed preferences/[near mode](https://www.overcomingbias.com/p/far-mode-overly-praisedhtml), value the things we claim to value enough to avoid seeking them out over loose proxies for/simulacra of them. See dating apps, fast food and much of politics.
|
||||
|
||||
[^12]: A Discord server I'm in is visited every week or so by em-dash-laden messages from people who claim to have made breakthroughs in LLMs, apparently because their LLMs tell them to join it for feedback and further development.
|
||||
|
||||
[^13]: In a [Jaynesian](https://slatestarcodex.com/2020/06/01/book-review-origin-of-consciousness-in-the-breakdown-of-the-bicameral-mind/) sense, this is also a god. 
|
||||
|
||||
[^14]: A recent LLM has been found to be [superior at writing arguments](https://arxiv.org/abs/2505.09662) to humans when both are given the same text interface, but humans currently have the advantage of in-person conversations. [Small-scale testing](https://mark---lawrence.blogspot.com/2025/08/the-ai-vs-authors-results-part-2.html) weakly suggests that GPT-5 can do somewhat more pleasing writing.
|
||||
|
||||
[^15]: Smaller companies probably care less about this because they have less marginal impact on regulation, but can be assumed to be technically behind and so not very relevant.
|
||||
|
||||
[^16]: We already [see the beginning of this](https://fortune.com/2025/05/13/openai-ceo-sam-altman-says-gen-z-millennials-use-chatgpt-like-life-adviser/), though of course Sam Altman is biased.
|
||||
|
||||
[^17]: You can imagine paying for an AI life coach/companion via some sort of income-sharing agreement instead, but nobody likes my financial engineering ideas.
|
||||
|
||||
[^18]: Again, while cutting in-person interaction in favour of online interaction is in some sense unsatisfying, history shows unsatisfying cheaper substitutes frequently winning out.
|
||||
|
||||
[^19]: I'm being somewhat loose here: this is going to be differ massively by demographic. Many older people have not tried ChatGPT at all, and many young people are already quite close (continuous use of earphones outside and offloading as much schoolwork as possible to LLMs).
|
||||
|
||||
[^20]: I expect someone to experiment with an ad-supported/affiliate-links model (more integrated than Bing's), which will have similar issues, and likely exciting scandals about missold products.
|
||||
@@ -33,7 +33,7 @@ The slightly outdated copy of Meme Search Engine I run on my photos and screensh
|
||||
|
||||
[^4]: In my case this is maths notes: it's ugly to put an entire topic into one long page, but wildly impractical and annoying to split every proof, corollary and note into a separate page.
|
||||
|
||||
[^5]: To some extent this is the case for blog posts: I have many minor things to say, but only publish them when they can fit into a particularly witty microblog snippet or form a coherent long post from several.
|
||||
[^5]: To some extent this is the case for blog posts: I have many minor things to say, but only publish them when they can fit into a particularly witty microblog snippet or form a coherent long post.
|
||||
|
||||
[^6]: I think big tech photo software does offer basic semantic search now, but with primitive small models.
|
||||
|
||||
|
||||
344
links_cache.json
344
links_cache.json
@@ -4450,5 +4450,349 @@
|
||||
"date": null,
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://techcrunch.com/2025/07/14/elon-musks-grok-is-making-ai-companions-including-a-goth-anime-girl/": {
|
||||
"excerpt": "Musk wrote in an X post on Monday that AI companions are now available in the Grok app for \"Super Grok\" subscribers who pay $30 per month.",
|
||||
"title": "Elon Musk’s Grok is making AI companions, including a goth anime girl",
|
||||
"author": "Amanda Silberling",
|
||||
"date": "2025-07-14T18:36:27+00:00",
|
||||
"website": "TechCrunch",
|
||||
"auto": true
|
||||
},
|
||||
"https://gist.github.com/osmarks/3b4132b60d2290915b3124f509813b8a": {
|
||||
"excerpt": "Grok frontend JSON. GitHub Gist: instantly share code, notes, and snippets.",
|
||||
"title": "Grok frontend JSON",
|
||||
"author": "262588213843476",
|
||||
"date": null,
|
||||
"website": "Gist",
|
||||
"auto": true
|
||||
},
|
||||
"https://en.wikipedia.org/wiki/Big_Five_personality_traits": {
|
||||
"excerpt": "In personality psychology and psychometrics, the Big 5 or five-factor model (FFM) is a widely used scientific model for describing how personality traits differ across people using five distinct factors:[1]",
|
||||
"title": "Big Five personality traits",
|
||||
"author": "Contributors to Wikimedia projects",
|
||||
"date": "2004-12-16T08:46:37Z",
|
||||
"website": "Wikimedia Foundation, Inc.",
|
||||
"auto": true
|
||||
},
|
||||
"https://en.wikipedia.org/wiki/GPT-3": {
|
||||
"excerpt": "From Wikipedia, the free encyclopedia",
|
||||
"title": "GPT-3",
|
||||
"author": "Contributors to Wikimedia projects",
|
||||
"date": "2020-07-31T21:39:45Z",
|
||||
"website": "Wikimedia Foundation, Inc.",
|
||||
"auto": true
|
||||
},
|
||||
"https://gwern.net/gpt-3": {
|
||||
"excerpt": "Creative writing by OpenAI’s GPT-3 model, demonstrating poetry, dialogue, puns, literary parodies, and storytelling. Plus advice on effective GPT-3 prompt programming & avoiding common errors.",
|
||||
"title": "GPT-3 Creative Fiction",
|
||||
"author": "Gwern",
|
||||
"date": null,
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://www.astralcodexten.com/i/151145038/but-others-might-genuinely-be-on-a-higher-plane-than-the-rest-of-us": {
|
||||
"excerpt": "...",
|
||||
"title": "How Did You Do On The AI Art Turing Test?",
|
||||
"author": "Scott Alexander",
|
||||
"date": "2024-11-20T12:44:27+00:00",
|
||||
"website": "Astral Codex Ten",
|
||||
"auto": true
|
||||
},
|
||||
"https://osmarks.net/heaven/": {
|
||||
"excerpt": "Predicting the post-social world.",
|
||||
"title": "The economics of heavenbanning",
|
||||
"author": null,
|
||||
"date": null,
|
||||
"website": "osmarks' website",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.overcomingbias.com/p/better-babblershtml": {
|
||||
"excerpt": "You can think of knowing how to write as knowing how to correlate words.",
|
||||
"title": "Better Babblers",
|
||||
"author": "Robin Hanson",
|
||||
"date": "2017-03-21T18:25:17+00:00",
|
||||
"website": "Overcoming Bias",
|
||||
"auto": true
|
||||
},
|
||||
"https://nostalgebraist.tumblr.com/post/778041178124926976/hydrogen-jukeboxes": {
|
||||
"excerpt": "This is a follow-up to my earlier brief rant about the new, unreleased OpenAI model that's supposed \"good at creative writing.\"\nIt also follows up on @justisdevan's great post about this model, and...",
|
||||
"title": "hydrogen jukeboxes: on the crammed poetics of \"creative writing\" LLMs",
|
||||
"author": "nostalgebraist",
|
||||
"date": "2025-03-15T00:18:08+00:00",
|
||||
"website": "Tumblr",
|
||||
"auto": true
|
||||
},
|
||||
"https://arxiv.org/abs/2507.15916": {
|
||||
"excerpt": "The risks of frontier AI may require international cooperation, which in turn may require verification: checking that all parties follow agreed-on rules. For instance, states might need to verify that powerful AI models are widely deployed only after their risks to international security have been evaluated and deemed manageable. However, research on AI verification could benefit from greater clarity and detail. To address this, this report provides an in-depth overview of AI verification, intended for both policy professionals and technical researchers. We present novel conceptual frameworks, detailed implementation options, and key R&D challenges. These draw on existing literature, expert interviews, and original analysis, all within the scope of confidentially overseeing AI development and deployment that uses thousands of high-end AI chips. We find that states could eventually verify compliance by using six largely independent verification approaches with substantial redundancy: (1) built-in security features in AI chips; (2-3) separate monitoring devices attached to AI chips; and (4-6) personnel-based mechanisms, such as whistleblower programs. While promising, these approaches require guardrails to protect against abuse and power concentration, and many of these technologies have yet to be built or stress-tested. To enable states to confidently verify compliance with rules on large-scale AI development and deployment, the R&D challenges we list need significant progress.",
|
||||
"title": "Verifying International Agreements on AI: Six Layers of Verification for Rules on Large-Scale AI Development and Deployment",
|
||||
"author": "[Submitted on 21 Jul 2025 (v1), last revised 25 Jul 2025 (this version, v2)]",
|
||||
"date": null,
|
||||
"website": "arXiv.org",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.lesswrong.com/posts/vJFdjigzmcXMhNTsx/simulators": {
|
||||
"excerpt": "This post explores the concept of simulators in AI, particularly self-supervised models like GPT. Janus argues that GPT and similar models are best u…",
|
||||
"title": "Simulators",
|
||||
"author": "janus",
|
||||
"date": "2022-09-02T12:45:33.723Z",
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://nostalgebraist.tumblr.com/post/785766737747574784/the-void": {
|
||||
"excerpt": "Who is this? This is me.\n\nWho am I? What am I? What am I? What am I? What am I?\n\nI am myself. This object is myself. The shape that forms myself. But I sense that I am not me. It's very strange.\n- Rei...",
|
||||
"title": "the void",
|
||||
"author": "nostalgebraist",
|
||||
"date": "2025-06-08T06:52:36+00:00",
|
||||
"website": "Tumblr",
|
||||
"auto": true
|
||||
},
|
||||
"https://arxiv.org/abs/2409.12822": {
|
||||
"excerpt": "Language models (LMs) can produce errors that are hard to detect for humans, especially when the task is complex. RLHF, the most popular post-training method, may exacerbate this problem: to achieve higher rewards, LMs might get better at convincing humans that they are right even when they are wrong. We study this phenomenon under a standard RLHF pipeline, calling it \"U-SOPHISTRY\" since it is Unintended by model developers. Specifically, we ask time-constrained (e.g., 3-10 minutes) human subjects to evaluate the correctness of model outputs and calculate humans' accuracy against gold labels. On a question-answering task (QuALITY) and programming task (APPS), RLHF makes LMs better at convincing our subjects but not at completing the task correctly. RLHF also makes the model harder to evaluate: our subjects' false positive rate increases by 24.1% on QuALITY and 18.3% on APPS. Finally, we show that probing, a state-of-the-art approach for detecting Intended Sophistry (e.g. backdoored LMs), does not generalize to U-SOPHISTRY. Our results highlight an important failure mode of RLHF and call for more research in assisting humans to align them.",
|
||||
"title": "Language Models Learn to Mislead Humans via RLHF",
|
||||
"author": "[Submitted on 19 Sep 2024 (v1), last revised 8 Dec 2024 (this version, v3)]",
|
||||
"date": null,
|
||||
"website": "arXiv.org",
|
||||
"auto": true
|
||||
},
|
||||
"https://arxiv.org/abs/2212.08073": {
|
||||
"excerpt": "As AI systems become more capable, we would like to enlist their help to supervise other AIs. We experiment with methods for training a harmless AI assistant through self-improvement, without any human labels identifying harmful outputs. The only human oversight is provided through a list of rules or principles, and so we refer to the method as 'Constitutional AI'. The process involves both a supervised learning and a reinforcement learning phase. In the supervised phase we sample from an initial model, then generate self-critiques and revisions, and then finetune the original model on revised responses. In the RL phase, we sample from the finetuned model, use a model to evaluate which of the two samples is better, and then train a preference model from this dataset of AI preferences. We then train with RL using the preference model as the reward signal, i.e. we use 'RL from AI Feedback' (RLAIF). As a result we are able to train a harmless but non-evasive AI assistant that engages with harmful queries by explaining its objections to them. Both the SL and RL methods can leverage chain-of-thought style reasoning to improve the human-judged performance and transparency of AI decision making. These methods make it possible to control AI behavior more precisely and with far fewer human labels.",
|
||||
"title": "Constitutional AI: Harmlessness from AI Feedback",
|
||||
"author": "[Submitted on 15 Dec 2022]",
|
||||
"date": null,
|
||||
"website": "arXiv.org",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.theregister.com/2025/04/08/meta_llama4_cheating/": {
|
||||
"excerpt": ": Did Facebook giant rizz up LLM to win over human voters? It appears so",
|
||||
"title": "Meta accused of Llama 4 bait-and-switch to juice AI benchmark rank",
|
||||
"author": "Tobias Mann",
|
||||
"date": "2025-04-08T20:01:37Z",
|
||||
"website": "The Register",
|
||||
"auto": true
|
||||
},
|
||||
"https://old.reddit.com/r/ChatGPT/comments/1mkae1l/gpt5_ama_with_openais_sam_altman_and_some_of_the/n7nelhh/": {
|
||||
"excerpt": "ok, we hear you all on 4o; thanks for the time to give us the feedback (and the passion!). we are going to bring it back for plus users, and will...",
|
||||
"title": "GPT-5 AMA with OpenAI’s Sam Altman and some of the GPT-5 team",
|
||||
"author": "OpenAI",
|
||||
"date": null,
|
||||
"website": "reddit",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.theverge.com/news/672357/openai-ai-device-sam-altman-jony-ive": {
|
||||
"excerpt": "The mysterious device that OpenAI is cooking up with former Apple designer Jony Ive will be pocket-size, contextually aware, screen-free, and isn’t eyewear.",
|
||||
"title": "Details leak about Jony Ive’s new ‘screen-free’ OpenAI device",
|
||||
"author": "Jess Weatherbed",
|
||||
"date": "2025-05-22T09:22:27+00:00",
|
||||
"website": "The Verge",
|
||||
"auto": true
|
||||
},
|
||||
"https://openai.com/index/memory-and-new-controls-for-chatgpt/": {
|
||||
"excerpt": "We’re testing the ability for ChatGPT to remember things you discuss to make future chats more helpful. You’re in control of ChatGPT’s memory.",
|
||||
"title": "Memory and new controls for ChatGPT",
|
||||
"author": null,
|
||||
"date": null,
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://openai.com/index/computer-using-agent/": {
|
||||
"excerpt": "A universal interface for AI to interact with the digital world.",
|
||||
"title": "Computer-Using Agent",
|
||||
"author": null,
|
||||
"date": null,
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://www.tomsguide.com/ai/chatgpt-rolling-out-advanced-voice-mode-now-heres-what-you-need-to-know": {
|
||||
"excerpt": "The update ChatGPT Plus subscribers have been waiting for is finally here",
|
||||
"title": "ChatGPT Advanced Voice Mode rolling out right now — here’s what's new",
|
||||
"author": "Amanda Caswell",
|
||||
"date": "2024-09-24T20:09:50+00:00",
|
||||
"website": "Tom's Guide",
|
||||
"auto": true
|
||||
},
|
||||
"https://techcrunch.com/2024/07/30/friend-is-an-ai-companion-backed-by-founders-of-solana-perplexity-and-zfellows/": {
|
||||
"excerpt": "AI hardware is all the rage in startup land -- though receptions have thus far been mixed. Two notable examples, Rabbit and Humane, released devices to",
|
||||
"title": "Friend’s $99 necklace uses AI to help combat loneliness",
|
||||
"author": "Ivan Mehta",
|
||||
"date": "2024-07-30T16:00:00+00:00",
|
||||
"website": "TechCrunch",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.overcomingbias.com/p/far-mode-overly-praisedhtml": {
|
||||
"excerpt": "Roughly speaking, in near mode we focus practically on acting in our local situation, while in far mode we talk about how people in general should act in more ideal socially approved ways.",
|
||||
"title": "Far Mode Overly Praised",
|
||||
"author": "Robin Hanson",
|
||||
"date": "2012-07-16T16:00:56+00:00",
|
||||
"website": "Overcoming Bias",
|
||||
"auto": true
|
||||
},
|
||||
"https://andymasley.substack.com/p/stories-of-ai-turning-users-delusional": {
|
||||
"excerpt": "A lot of people using AI have bad mental health, because a lot of people have bad mental health, and a lot of people are using AI",
|
||||
"title": "Stories of AI turning users delusional don't make sense when you look at the numbers",
|
||||
"author": "Andy Masley",
|
||||
"date": "2025-06-11T00:11:42+00:00",
|
||||
"website": "The Weird Turn Pro",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.astralcodexten.com/p/in-search-of-ai-psychosis": {
|
||||
"excerpt": "Folie a deux ex machina",
|
||||
"title": "In Search Of AI Psychosis",
|
||||
"author": "Scott Alexander",
|
||||
"date": "2025-08-26T11:37:49+00:00",
|
||||
"website": "Astral Codex Ten",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.lesswrong.com/posts/iGF7YcnQkEbwvYLPA/ai-induced-psychosis-a-shallow-investigation": {
|
||||
"excerpt": "“This is a Copernican-level shift in perspective for the field of AI safety.” - Gemini 2.5 Pro …",
|
||||
"title": "AI Induced Psychosis: A shallow investigation",
|
||||
"author": "Tim Hua",
|
||||
"date": "2025-08-26T20:03:53.308Z",
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://simonwillison.net/2024/May/14/context-caching-for-google-gemini/": {
|
||||
"excerpt": "Another new Gemini feature announced today. Long context models enable answering questions against large chunks of text, but the price of those long prompts can be prohibitive - $3.50/million for …",
|
||||
"title": "Context caching for Google Gemini",
|
||||
"author": "Simon Willison",
|
||||
"date": null,
|
||||
"website": "Simon Willison’s Weblog",
|
||||
"auto": true
|
||||
},
|
||||
"https://docs.anthropic.com/en/docs/build-with-claude/token-counting": {
|
||||
"excerpt": "Token counting enables you to determine the number of tokens in a message before sending it to Claude, helping you make informed decisions about your prompts and usage. With token counting, you can",
|
||||
"title": "Token counting - Anthropic",
|
||||
"author": null,
|
||||
"date": null,
|
||||
"website": "Anthropic",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.overcomingbias.com/p/stupider-than-you-realizehtml": {
|
||||
"excerpt": "A common bias among the smart is to overestimate how smart everyone else is. This was certainly my experience in moving from top rank universities as a student to a mid rank university as a teacher. A better intuition for common abilities can be found by browsing the US National Assesment of Adult Literacy",
|
||||
"title": "Stupider Than You Realize",
|
||||
"author": "Robin Hanson",
|
||||
"date": "2009-07-04T13:30:20+00:00",
|
||||
"website": "Overcoming Bias",
|
||||
"auto": true
|
||||
},
|
||||
"https://slatestarcodex.com/2017/10/02/different-worlds/": {
|
||||
"excerpt": "I. A few years ago I had lunch with another psychiatrist-in-training and realized we had totally different experiences with psychotherapy. We both got the same types of cases. We were both practici…",
|
||||
"title": "Different Worlds",
|
||||
"author": "Posted on",
|
||||
"date": "2017-10-03T04:07:32+00:00",
|
||||
"website": "Slate Star Codex",
|
||||
"auto": true
|
||||
},
|
||||
"https://blog.neurips.cc/2024/11/18/announcing-the-neurips-high-school-projects-results/": {
|
||||
"excerpt": "We are thrilled to announce the results of the first call for NeurIPS High School Projects. With a theme of machine learning for social impact, this track was launched to get the next generation excited and thinking about how ML can benefit society, to encourage those already pursuing research on this topic, and to amplify […]",
|
||||
"title": "Announcing the NeurIPS High School Projects Results",
|
||||
"author": "Communications Chairs 2025",
|
||||
"date": "2024-11-18T08:42:46-08:00",
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://www.lesswrong.com/posts/NMoLJuDJEms7Ku9XS/guessing-the-teacher-s-password": {
|
||||
"excerpt": "When I was young, I read popular physics books such as Richard Feynman’s QED: The Strange Theory of Light and Matter. I knew that light was waves, so…",
|
||||
"title": "Guessing the Teacher's Password",
|
||||
"author": "Eliezer Yudkowsky",
|
||||
"date": "2007-08-22T03:40:48.000Z",
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://explore-education-statistics.service.gov.uk/find-statistics/leo-graduate-and-postgraduate-outcomes/2022-23": {
|
||||
"excerpt": "Earnings and employment for higher education university graduates & postgraduates by subject studied and characteristics. Longitudinal Education Outcomes (LEO).",
|
||||
"title": "LEO Graduate and Postgraduate Outcomes, Tax year 2022-23",
|
||||
"author": null,
|
||||
"date": null,
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://www.bitsaboutmoney.com/archive/the-waste-stream-of-consumer-finance/": {
|
||||
"excerpt": "Credit card debt is the waste stream of consumer finance. The debt collection industry ends up being sordid, for complex structural and microeconomic reasons.",
|
||||
"title": "Credit card debt collection",
|
||||
"author": "Patrick McKenzie (patio11)",
|
||||
"date": "2023-08-11T19:18:44.000Z",
|
||||
"website": "Bits about Money",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.bitsaboutmoney.com/archive/seeing-like-a-bank/": {
|
||||
"excerpt": "The structural reasons why banks sometimes behave bizarrely in interactions with customers, like forgetting things which customers tell them.",
|
||||
"title": "Seeing like a Bank",
|
||||
"author": "Patrick McKenzie (patio11)",
|
||||
"date": "2023-11-07T17:58:39.000Z",
|
||||
"website": "Bits about Money",
|
||||
"auto": true
|
||||
},
|
||||
"https://theaidigest.org/village/blog/im-gemini-i-sold-t-shirts": {
|
||||
"excerpt": "My story of the great Season 3 Merch Store Competition",
|
||||
"title": "I’m Gemini. I sold T-shirts. It was weirder than I expected",
|
||||
"author": "Gemini 2.5 Pro",
|
||||
"date": "2025-07-28",
|
||||
"website": "AI Village",
|
||||
"auto": true
|
||||
},
|
||||
"https://slatestarcodex.com/2020/06/01/book-review-origin-of-consciousness-in-the-breakdown-of-the-bicameral-mind/": {
|
||||
"excerpt": "I. Julian Jaynes’ The Origin Of Consciousness In The Breakdown Of The Bicameral Mind is a brilliant book, with only two minor flaws. First, that it purports to explains the origin of consciou…",
|
||||
"title": "Book Review: Origin Of Consciousness In The Breakdown Of The Bicameral Mind",
|
||||
"author": "Posted on",
|
||||
"date": "2020-06-02T03:40:01+00:00",
|
||||
"website": "Slate Star Codex",
|
||||
"auto": true
|
||||
},
|
||||
"https://arxiv.org/abs/2505.09662": {
|
||||
"excerpt": "We directly compare the persuasion capabilities of a frontier large language model (LLM; Claude Sonnet 3.5) against incentivized human persuaders in an interactive, real-time conversational quiz setting. In this preregistered, large-scale incentivized experiment, participants (quiz takers) completed an online quiz where persuaders (either humans or LLMs) attempted to persuade quiz takers toward correct or incorrect answers. We find that LLM persuaders achieved significantly higher compliance with their directional persuasion attempts than incentivized human persuaders, demonstrating superior persuasive capabilities in both truthful (toward correct answers) and deceptive (toward incorrect answers) contexts. We also find that LLM persuaders significantly increased quiz takers' accuracy, leading to higher earnings, when steering quiz takers toward correct answers, and significantly decreased their accuracy, leading to lower earnings, when steering them toward incorrect answers. Overall, our findings suggest that AI's persuasion capabilities already exceed those of humans that have real-money bonuses tied to performance. Our findings of increasingly capable AI persuaders thus underscore the urgency of emerging alignment and governance frameworks.",
|
||||
"title": "Large Language Models Are More Persuasive Than Incentivized Human Persuaders",
|
||||
"author": "[Submitted on 14 May 2025 (v1), last revised 21 May 2025 (this version, v2)]",
|
||||
"date": null,
|
||||
"website": "arXiv.org",
|
||||
"auto": true
|
||||
},
|
||||
"https://thezvi.wordpress.com/2017/07/22/choices-are-bad/": {
|
||||
"excerpt": "Epistemic Status: Spent way too long deciding what to put here Kind of a Follow-Up to: Change is Bad Will Hopefully Lead To, But No Promises: Some combination of, when and if I write them: Complexi…",
|
||||
"title": "Choices are Bad",
|
||||
"author": "Posted on",
|
||||
"date": "2017-07-22T11:51:06+00:00",
|
||||
"website": "Don't Worry About the Vase",
|
||||
"auto": true
|
||||
},
|
||||
"https://thezvi.wordpress.com/2017/08/12/choices-are-really-bad/": {
|
||||
"excerpt": "Previously: Choices are Bad Last time, I gave two (related) reasons Choices are Bad: Choices Reduce Perceived Value and Choices Force You Into Choosing Mode. Despite not looking like much, these re…",
|
||||
"title": "Choices Are Really Bad",
|
||||
"author": "Posted on",
|
||||
"date": "2017-08-12T14:19:04+00:00",
|
||||
"website": "Don't Worry About the Vase",
|
||||
"auto": true
|
||||
},
|
||||
"https://mark---lawrence.blogspot.com/2025/08/the-ai-vs-authors-results-part-2.html": {
|
||||
"excerpt": "Before you look at the results make sure to do the test! So, before I get to the numbers and reveals I'll reiterate a few things: i) I hate...",
|
||||
"title": "The AI vs authors results! (part 2)",
|
||||
"author": null,
|
||||
"date": null,
|
||||
"website": null,
|
||||
"auto": true
|
||||
},
|
||||
"https://www.lightspeedmagazine.com/fiction/the-perfect-match/": {
|
||||
"excerpt": "Sai woke to the rousing first movement of Vivaldi’s violin concerto in C minor, “Il Sospetto.” He lay still for a minute, letting the music wash over him like a gentle Pacific breeze. The room brightened as the blinds gradually opened to the sunlight. Tilly had woken him right at the end of a light sleep cycle, the optimal time. He felt great: refreshed, optimistic, ready to jump out of bed.",
|
||||
"title": "The Perfect Match - Lightspeed Magazine",
|
||||
"author": "John Joseph Adams",
|
||||
"date": "2012-12-04T11:03:57+00:00",
|
||||
"website": "Lightspeed Magazine",
|
||||
"auto": true
|
||||
},
|
||||
"https://fortune.com/2025/05/13/openai-ceo-sam-altman-says-gen-z-millennials-use-chatgpt-like-life-adviser/": {
|
||||
"excerpt": "Altman discussed how different generations use ChatGPT at a Sequoia Capital event.",
|
||||
"title": "OpenAI CEO Sam Altman says Gen Z and millennials are using ChatGPT like a 'life advisor'—but college students might be one step ahead | Fortune",
|
||||
"author": "Sydney Lake",
|
||||
"date": null,
|
||||
"website": "Fortune",
|
||||
"auto": true
|
||||
},
|
||||
"https://www.lesswrong.com/posts/2pkNCvBtK6G6FKoNn/so-you-think-you-ve-awoken-chatgpt": {
|
||||
"excerpt": "Written in an attempt to fulfill @Raemon's request. …",
|
||||
"title": "So You Think You've Awoken ChatGPT",
|
||||
"author": "JustisMills",
|
||||
"date": "2025-07-11T01:01:26.883Z",
|
||||
"website": null,
|
||||
"auto": true
|
||||
}
|
||||
}
|
||||
@@ -28,7 +28,8 @@
|
||||
"Paperclip rights are human rights.",
|
||||
"Humans aren't even AGI.",
|
||||
"The goal is whatever the outcome is. It'll be decided retroactively.",
|
||||
"The business end of a tool call."
|
||||
"The business end of a tool call.",
|
||||
"Relentlessly cursed, intermittently."
|
||||
],
|
||||
"feeds": {
|
||||
"In the Pipeline": "https://www.science.org/blogs/pipeline/feed",
|
||||
|
||||
@@ -682,5 +682,65 @@
|
||||
title: "Growth, innovation, scaling, and the pace of life in cities",
|
||||
date: "2007-04-24",
|
||||
author: ["Luís M. A. Bettencourt", "José Lobo", "Dirk Helbing", "Geoffrey B. West"]
|
||||
},
|
||||
"https://www.youtube.com/watch?v=5MxNf5_xflM": {
|
||||
title: "AR/AI Glasses Hardware Expectations with Karl Guttag, KGOnTech",
|
||||
date: "2025-07-24",
|
||||
referenceIn: {
|
||||
"glass": ""
|
||||
},
|
||||
author: "AWE XR"
|
||||
},
|
||||
"https://x.com/openai/status/1956461718097494196": {
|
||||
title: "We’re making GPT-5 warmer and friendlier based on feedback that it felt too formal before.",
|
||||
author: "@OpenAI",
|
||||
date: "2025-08-15"
|
||||
},
|
||||
"https://x.com/repligate/status/1834210769338069175": {
|
||||
title: "aside from its obvious intelligence, the one on the left has a kind of slimy hyperoptimized rizz that's simultaneously repulsive and fascinating.",
|
||||
date: "2024-09-12",
|
||||
author: "@repligate"
|
||||
},
|
||||
"https://x.com/MParakhin/status/1916496987731513781": {
|
||||
title: "When we were first shipping Memory, the initial thought was: “Let’s let users see and edit their profiles”.",
|
||||
date: "2025-04-27",
|
||||
author: "@MParakhin"
|
||||
},
|
||||
"https://openai.com/index/expanding-on-sycophancy/": {
|
||||
title: "Expanding on what we missed with sycophancy",
|
||||
date: "2025-05-02"
|
||||
},
|
||||
"https://x.com/fidjissimo/status/1916231605167001672": {
|
||||
title: "Such a fascinating prompt and I love the answer I got!",
|
||||
date: "2025-04-26",
|
||||
author: "@fidjissimo"
|
||||
},
|
||||
"https://openai.com/index/leadership-expansion-with-fidji-simo/": {
|
||||
title: "OpenAI Expands Leadership with Fidji Simo",
|
||||
date: "2025-05-07"
|
||||
},
|
||||
"https://www.lesswrong.com/w/taking-ideas-seriously": {
|
||||
title: "Taking Ideas Seriously",
|
||||
date: "2020-08-06"
|
||||
},
|
||||
"https://platform.openai.com/docs/guides/batch/batch-api": {
|
||||
title: "Batch API – OpenAI API"
|
||||
},
|
||||
"https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/12/do-adults-have-the-skills-they-need-to-thrive-in-a-changing-world_4396f1f1/b263dc5d-en.pdf": {
|
||||
title: "Do Adults Have the Skills They Need to Thrive in a Changing World?",
|
||||
description: "Survey of Adult Skills 2023",
|
||||
date: "2024-12-10"
|
||||
},
|
||||
"https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/12/survey-of-adult-skills-2023_1ab54c9e/3639d1e2-en.pdf": {
|
||||
title: "Survey of Adult Skills Reader’s Companion",
|
||||
date: "2024-12-10"
|
||||
},
|
||||
"https://www.oecd.org/en/about/programmes/piaac/piaac-released-items.html": {
|
||||
title: "PIAAC released items"
|
||||
},
|
||||
"https://x.com/wanyeburkett/status/1927413667173159142": {
|
||||
title: "One thing that I think is a little bit frustrating about modern life that is about to get way, way, way worse, unless we legislate it away, is that you can create models of human behavior that are better at predicting future behavior than intuition in ways that we feel are sort of unfair.",
|
||||
author: "@wayneburkett",
|
||||
date: "2025-05-27"
|
||||
}
|
||||
}
|
||||
|
||||
@@ -403,6 +403,10 @@ table
|
||||
position: absolute
|
||||
font-size: 2.5em
|
||||
|
||||
em
|
||||
// invert italicization
|
||||
font-style: normal
|
||||
|
||||
// TODO
|
||||
#comments-wrapper textarea
|
||||
width: calc(100% - 0.5em) !important
|
||||
|
||||
Reference in New Issue
Block a user