davidthewatson

No problem on the book suggestion.

Chutkan deserves the success she's had since that book shipped. The book is not an easy read. It requires a strong stomach, to say the least.

I was lucky to have a doctor who was thoughtful enough to connect my obsessions with diet and exercise to continuous improvement by leveraging my hyperlexia against the root cause that I have a stomach in each leg. ;)

I like your theory. It would be interesting to see a group of biology students debate the merits of such a theory and how they might prove or disprove such a theory.

I'm sure there are some game theory problems lurking there.

This resonates given my love of dogs and the huge chain of complex events that likely connects the sewage plant effluent that fertilized the lawn where I lived as a child and my emergent disease states that followed.

Ignoring the fact that exposure may be a problem from E. coli to sewage, I read Robynne Chutkan's microbiome book at the suggestion of my doctor several years ago and nothing has come closer to the truth on our lack of exposure also being part of the problem.

You're damned if you do and damned if you don't. Try finding the median there and there's a Nobel waiting for you.

https://robynnechutkan.com/

To say I'm speechless is a mild understatement.

Thanks so much for sharing upset given the context of my question!

Their explication of Venn diagrams is exemplary and compelling, particularly the banana and the way the Venn is elucidated and visualized through the accompanying meta-detail, for lack of a better phrase.

I cannot thank you enough. This will literally propel much of my work on data going forward.

Nice! Thanks.

Has anyone used loop? Radial menu with keyboard trigger caught my eye.

This makes sense!

However, as I've commented on other software like this, why are none of the examples style-free Venn diagrams? That is, black on white 3 circle overlapping each other in pairs with a single shared triangle intersection at center? I apologize if one of these is already present. Style has little value there since it may obscure reality.

This has already been posted but I'm afraid the title may have kept it low in the SEO race that is hacker news.

Of note:

    "Although the results from vanilla prompts and the user study build the illusion of ToM abilities in LLMs, we conclude that our perturbation experiments are sufficient to clarify that ToM abilities are absent from LLMs. ToM evaluation must entail evaluation of reasoning abilities with minimal leniency towards failure cases."

It's worth noting that the word "idempotent" does not occur anywhere in the text.

It seems like a latent conversation given how ToM has impacted everything from Autism research to consciousness research.

I may have missed the deeper discussion in my cursory reading, which is lossy and poor. If so, I'm sorry.

1) The idea that there are extant bidirectional games, whether intentional or unintentional, is absent, whether elicited by the LLM or the human interlocutor in conversation. These are best labeled as mentalism: the creation of illusions as a technique to delude the human or bypass the conditional guardrails of the LLM.

2) The introduction of bicameralism (Jaynes) along with ideas stemming from the "controlled hallucination" theory of Anil Seth and its further explication by Andy Clark is also absent.

I'd be happy if anyone has insight into these areas beyond Peter Pirolli's work:

https://www.efsa.europa.eu/sites/default/files/event/180918-...

on Information Foraging and so on.

It's interesting that the paper mentions Andy Clark's work and hallucinations in separate threads but doesn't connect Anil Seth's introduction of controlled hallucinations and Andy Clark's work on that topic:

https://www.technologyreview.com/2021/08/25/1032121/brains-c...

I had always conceived of metacognition as genetic, given my own phenomenological experience dating to childhood. It wasn't until a colleague challenged this idea a few years ago at Carnegie Mellon that I'd even considered that metacognition may be taught and learned.

I realized that I'd finally accepted my habit of reading Ken Wilbur and others sideways an odd paragraph at a time when I saw your comment here. Thank you for the idea that my nonlinearity is ok with a tip of the hat to leo buscaglia.

That's right. It's been a problem in war for decades.

I'm happy to see a return to the defensive survival evacuation and triage role for AI, CV, and robotics as opposed to the regression from that role into autonomous killing machines.

Building an Alan Alda MASH persona along with supporting performers and portrayers is not a bad idea, even if the dragging on the sled requires much less intelligence save for the head-banging of the injured that's tough to avoid and manage.

Technical innovation in metallurgy plays a significant role. It's easy to say that things are a distraction, but it's also only true in retrospect, unless you're on a spectrum of Turing award winners or the like.

It's not clear to me whether you meant "pivot" or "pilot" but the downstream effect is likely the same. Trying to account for divestitures and acquisitions in such a large corporation is a fool's gold of forensic accounting. Just because the business wasn't more transparent than their incorporation required doesn't mean that they weren't thinking broadly and deeply about the future of their core business.

Thanks for your response.

There was plenty of innovation clear to anyone with the proximity and education to recognize it, but your questions raise intractable issues of the proprietary information of any large corporation.

From my naive college kid's vantage point, the majority of innovation was tactical and not tied to core competencies owing to the siloed nature of communication between management and technical layers; i.e. it's difficult, if not impossible, for innovation to pass through that gated silo without a catalyst to nurture and support its transformation from tactical to strategic. This is a fundamental problem of complementary assets and the need for interdisciplinary culture that accrues to their composition, complexity, and collaborative community.

We could certainly debate whether this is the crux of "The Corporation" over the last twenty years of acquisitions seemingly aimed at silencing competition.

Of the two companies I worked with in that time period thirty years ago, both of their nail-in-the-coffin, final acquisitions approached $20 billion. Innovation has a way of disappearing in the dark subtext of large corporate acquisitions.

The "never given a chance to thrive" is insightful beyond US Steel. That is, well into modern tech startups that are managed by the same kinds of people who managed GM in the 1970's.

I've seen a lot of bad marketing copy, but this deserves to be placed on a pedestal of achievement for its tongue in cheek projection of how seriously they took the creation of such a funny thing:

"The Fart Machine is not just any ordinary guitar pedal or plug-in. Using our unique rectal microphone technology, we have modeled every component of the human bowel system, and now the Fart Machine can turn any sound into flatulence. This technology provides an unprecedented level of control over your fart sounds, enabling you to create a range of unique and unforgettable sounds that will leave your audience in awe."

I miss music software and its sampling and remixing culture.

I was asked to share a TA role on a graduate course in HPC a decade ago. I turned down the offer.

After a cursory glance, I can honestly say that if this book were available then, I'd have taken the opportunity.

The combination of what I perceive to be Knuth's framing of art, along with carpentry and the need to be a better devops person than your devops person is compelling.

Kudos to the author for such an achievement. UT Austin seems to have achieved in computer science what North Texas State did in music.

I'm happy to see the link here as I was curious about the subject given that my career started in US Steel's data center in Pittsburgh.

I'm saddened by the fact that this retelling seems mostly negative and ignores a large part of US Steel's evolution into USX. The retelling is subtractive whether you view Marathon Oil's involvement as a positive complementary asset play at the time or a negative given the history of its divestiture.

I can say that there was innovation in the data center where I worked in the evolution from manual human mainframe era data center operations to token ring networks of PC API's along with abstraction and automation via glue code.

The minimization of manual human labor as people retired is likely lost to the history books unless one of my old technical collaborators decides to write a book in retirement.

The CMU kids I worked with at US Steel's data center in Pittsburgh were just as smart as the ones I worked with in the software industry from Boston to Seattle.

The problem is that some medical devices are approved despite their being demonstrably deadly as they are designed, manufactured, shipped, and used without sufficient oversight by the governing body of apologists put in place to oversee them. This pattern has only worsened since therac 25.

This is amazing!

I'm saying that as a long-time user of a desktop setup consisting of Lin-Mac-Win boxes arranged left-to-right for at least two decades. Polyglot is to programming languages as multi-platform is to hardware. Fortunately, that's gotten easier as system-on-a-chip or SoC designs have proliferated in the last few years.

My current setup consists of 2x $250 Ryzen SoC Mac-mini copies straight outta Shenzhen and an M1 MBP.

I only wish that the once resilient world of open source KVM solutions such as synergy had not devolved into a recursive nightmare of forks where getting a working multi-platform open source software KVM without cost is challenging. Sadly, my current setup uses a good 'ole KVM switch but the button pressing is distracting.

I look forward to trying something like this again one day. Thanks for sharing!

Does anyone know if the libs discussed here can do a basic 3 circle Venn diagram with the intersection highlighted? I didn't see any Venn diagrams in the examples.

Thanks!

What founders, managers, and engineers share is the team responsibility for founding, managing, and engineering the product and its value proposition. A product's value proposition may be mostly subjective but is easily made objective by cost/revenue ratio.

Me too, especially when the underlying root cause is siloed communication arising from founder burnout or sheer disinterest not the supply of product management thinking. It's as if you needed the subtractive design thinking of the two Bobs in office space and what you got instead was aimless growth of product like cancer.

PDF shares a common property with computational complexity and robotics: the magical world of software is no longer free from physics, as if it ever was. Software has reciprocity in creating these illusions and destroying them.

I thought being emancipated from word docs was freedom until I realized the suffering brought on by a left field version of wkhtmltopdf no one uses but someone built and distributed via pamac AUR. In software, Hybridization is Postmodernism.

The paradox is that hybrids have MORE potential trouble areas than ICE but less reported trouble.

All hail McLuhan.

I have a Brother business-class (whatever that means) color laser printer sitting in my wife's home office. It's so perfect, it's beyond even the legendary HP Laserjet 4 it replaced when the old reliable fell over six feet from a shelf and went "ouch".

Brother deserves credit for its practice of stealth postmodern retrocomputing, for lack of a better phrase, if you will.

Cheap components are durable and scale, whether hardware, software, or SoC blurring the lines between the two.

Sadly, what this demonstrates is just how much the cloud dreams of Brother's competitors have diminishing returns because they are so tied up with authoritarian control regimes that have nothing to do with users' experience with their respective devices. Quite the converse.

The most interesting thing here is not the cult of personality battle between board and CEO. Rather, it's that these teams have managed to ship consumer AI that has a liminal, asymptotic edge where the smart kids can manipulate it into doing emergent things that it was not designed to do. That is, many of the outcomes of in-context learning could not be predicted at design time and they are, in fact, mind-blowing, magical, and likely not safe for consumption by those who believe that the machines are anywhere near the spectrum from consciousness to sentience.

Where do these fall with respect to the AI test generation tool I saw announced here from a YC-backed startup founder the other day and now can't find anywhere. I know selenium from a decade ago, but the others seem more emergent. Given my history metaprogramming tests when you have > 1 million LOC and zero tests, I'm hopeful that AI can actually implement the magic better than I can metaprogram it myself, i.e. there's a lot of technical debt in the metaprogramming approach since it's essentially unreadable gibberish that must be effectively loop-unrolled eventually into good, old-fashioned code that humans can actually read as opposed to rando collections of list comprehensions that look like the work of an insane lisp hacker.

My observations:

6 of the top 10 are rust. 4 of the top 10 are postgres.

Postgres needs little explanation given its history and Stonebraker's reputation.

I now understand why teams I've occupied had the strange loop crowd resigning to go work on rust.

Because Postgres has been profiled and tuned for performance for decades?

All hail Michael Stonebraker and Tom Lane. The Turing award and Wikipedia cannot be wrong.

Why?

I'd give more time to groceries if the experience of gathering them resulted in better nutrition, fun, and groceries existed in a third place that I really enjoyed like the old working space I occupied for a few years.

I guess the subtext is that I improvise groceries. I don't orchestrate them. I suspect consumer personas pivot on this inflection point around groceries. It's worth studying DFW "This is water" for context.

I was fortunate enough to see and hear Randy Pausch describe Alice's development in a lecture at CMU while he was still alive. I remember it being quite compelling as he discussed teaching young kids concurrency by animating a figure skater pirouetting on screen and the fact that the language made this straightforward compared to say, how this would be implemented in C++, for instance.

The problem, of course, is that the very affordance in a language like Alice that makes it approachable for novices are exactly the same constraints and limitations that drive experts insane.

The following is worth noting from the Wikipedia link you included:

"Alice is designed to appeal to specific subpopulations not normally exposed to computer programming, such as students of middle school age, by encouraging storytelling. Alice is also used at many colleges and universities in Introduction to Programming courses."

My point is simply that Alice seems like the wrong tool for the wrong age group as described by your experience but it's definitely not the right approach for someone who will major in computer science later.