This is a _dense_ book (in the best way). So much incredible history that's not as well presented anywhere else I've found. Sitting at 500+ pages and the font is pretty damn small, this is one hell of an informative book. We really stand on the shoulders of giants.
'The Dream Machine' languished in my 'To read' pile until I started the audiobook while on a road trip earlier this year. I was drawn in and read it through, switching between audio and ebook. I knew parts of the puzzle but it was amazing to see how they all fit together. An amazing book about a series of amazing accomplishments that let us share these ideas so widely and quickly.
The Stripe Press edition of the book is a beautiful product, but the font is small. I bought the book but had a hard time reading it so I got a digital copy for my ereader where I could select a larger font.
It is one of my favorites. I actually came to it via "where wizards stay up late", which I would also recommend although it doesn't have the depth and insight of dream machine.
Wizards for me falls flat (kinda like "soul of a new machine" that's often recommended but which I found a complete waste of time) while the Dream Machine, Dealers of Lightning [1] and Norbert Wiener's biography [2] are all essential reading.
"The Soul of a New Machine" is not so much a foundational book but a snapshot of the computer industry and might be more interesting to someone who had no idea of the industry as I did when it was published in 1981. For me it was a fascinating glimpse into the tech industry that I would soon join, albeit in software not hardware.
Yeah. I think if I read Dream Machine first I would have enjoyed Wizards less. But some of the history was new to me which meant that I enjoyed Wizards for those aspects and then Dream Machine for the detailed look into why it happened.
1. Not to be confused with Ted Nelson's titled work Dream Machines, one half of the pair Computer Lib/Dream Machines that described Project Xanadu and articulated a vision about computer user freedom that predates Stallman and the free software movement, complete with manifesto.
In humans, the cerebral cortex carries out functions, like memory, thinking, learning, reasoning, problem-solving, emotions, consciousness, and sensory functions.
Impressively, the author predicts the rise of declarative programming:
> "The idea may be highlighted by comparing instructions ordinarily addressed to intelligent human beings with instructions ordinarily used with computers. The latter specify precisely the individual steps to take and the sequence in which to take them. The former present or imply something about incentive or motivation, and they supply a criterion by which the human executor of the instructions will know when he has accomplished his task. In short: instructions directed to computers specify courses; instructions-directed to human beings specify goals. ... It is clear that, working within the loose constraints of predetermined strategies, computers will in due course be able to devise and simplify their own procedures for achieving stated goals."
I was going to quip that "idea guys" can predict things in a useless way when having no idea how to make them a reality...but Licklider went on to have such impact on our field.
It's mindblowing he was central to the creation of Arpanet and multics and so many other critical foundational bricks of today's computing.
> The fig tree is pollinated only by the insect Blastophaga grossorun. The larva of the insect lives in the ovary of the fig tree, and there it gets its food. The tree and the insect are thus heavily interdependent: the tree cannot reproduce without the insect; the insect cannot eat without the tree; together, they constitute not only a viable but a productive and thriving partnership. This cooperative "living together in intimate association, or even close union, of two dissimilar organisms" is called symbiosis [27].
I believe the meaning of symbiosis has evolved since 1960. Symbiosis is now an umbrella term that only implies a close association between two organisms. It doesn't necessarily imply mutual benefit anymore. A mutually beneficial association is called mutualism. An association where one benefits at the expense of the other is parasitism. Commensalism is when one benefits and the other is neither harmed nor benefited. There is also amensalism, where one is harmed and the other is neither harmed nor benefited. Symbiosis is the superset of parasitism, commensalism, amensalism, and mutualism.
It's amusing to me when people use "symbiosis" in the old sense and then I start pondering whether the thing they're describing is actually {mutual,amensal,parasit,commensal}ism
I think you're confusing a technical definition used among certain academic milieus with an everyday meaning of the rest of a language's speakers whose meaning has not changed since the 1960s. If someone uses the word symbiosis without referring to the distinctions you mention, you can just assume they're referring to what you call mutualism. We do this all the time in language. If someone says they're going to check the internet for something, I don't say well actually what you mean is you're launching your browser and only accessing a subset of the internet called the web... Of course not.
I get it. The everyday meaning is mutualism. I'm not confusing anything; please don't write off my perspective as mere "confusion". We are on HN, a place founded on gratifying curiosity. I was fascinated to learn the broader meaning of the term and figured that HN community would also enjoy discussing the rich technical meaning of the term. But instead I got some amusing interactions with the Semantics Police. And it's not like I'm standing on shaky ground with my definition; as I've mentioned before I am sourcing it from the first paragraph of the Wikipedia article on symbiosis. But yes, re-reading my original comment, I can see why you all might say "no, YOU're trying to be the Semantics Police!" Not my intention. I can easily hold both meanings of the term in my mind at the same time. The thing I really wanted to focus on was the fascinating richer meaning of "symbiosis" that makes space for benefit/benefit, harm/benefit, benefit/neutral, harm/neutral interactions.
The only thing I take back is the phrasing about it "evolving" since the 1960s. I looked up the biological terms; mutualism goes back to the 1840s at least. I guess Proudhon coined it. Amensalism goes back to the 1870s. Etc. It's still plausible to me that biologists or ecologists really pinned down how these terms relate to each other in the 1960s or later (since ecology as a field of study really ramped up in the 60s) but that would be difficult to prove.
Also, please remember that the topic came up because Licklider, an academic computer scientist and psychologist, used a biological term loosely in the title of an academic article published to an official computer science journal.
Much of what you write is interesting, and I appreciate your reflection. I'd just like to qualify something about Licklider. At the time of the Man-Computer Symbiosis in 1960 computer science didn't exist as a discipline, so he wasn't writing as a computer scientist. In fact, Licklider was instrumental in the establishment of computer science as a discipline through directing funds to establish the first departments for it in the US later in the '60s. (I have 1963 in my head but could be wrong.)
A more apt descriptor for him would be cybernetician. He was an acolyte of Norbert Wiener and his man-computer symbiosis can and should be read as his attempt to resolve the conflict in Wiener's writings between humans and machines. Instead of being antagonists, they could mutually benefit from one another as partners in a symbiosis.
FWIW, I’ve never heard of any meaning of symbiosis other than mutually beneficial. I’m not able to cite this but I believe the prefix sym- may imply this.
Look at the first paragraph of the Wikipedia page on the topic
> Symbiosis (from Greek συμβίωσις, symbíōsis, "living with, companionship, camaraderie", from σύν, sýn, "together", and βίωσις, bíōsis, "living")[2] is any type of a close and long-term biological interaction between two biological organisms of different species, termed symbionts, be it mutualistic, commensalistic, or parasitic.[3] In 1879, Heinrich Anton de Bary defined it as "the living together of unlike organisms". The term is sometimes used in the more restricted sense of a mutually beneficial interaction in which both symbionts contribute to each other's support.[3]
It acknowledges the popular use of the term while implying that the technical meaning is broader
> I’ve never heard of any meaning of symbiosis other than mutually beneficial
This is exactly why it's so funny to me. Some snake oil salesman says "we have a symbiotic relationship with XYZ" and in my head I'm thinking "yes, a parasitic symbiotic relationship..."
Google Trends for parasitic symbiosis etal seems interestingly different for US[1] and Worldwide[2], suggesting the possibility of regionally divergent usage.
It's all fun and games by Lewis Carroll, but given the amount of doublespeak in American cities these days (especially by the Left but also by the Right), I think we need to insist that words have a meaning and you can't twist them to mean the opposite.
I am literally sourcing my definition from the first paragraph of the Wikipedia article on symbiosis
I totally get it. People usually mean "mutually beneficial" when they talk about symbiosis. But the deeper we go into this Hacker News thread, the more amusing it gets to me that multiple commenters are really struggling to acknowledge the broader meaning of the term. And what's even funnier is that the broader meaning probably originates from biologists and ecologists; people who are very invested in precise language in this space.
[EDIT] Apologies for the formatting, iOS OCR isn’t the greatest and neither is HN block-quote formatting (or I just don’t know how to use it correctly.)
Now now, we all know we can’t just cite Wikipedia as a source. I hardly think this is as settled a matter as you and the author of the article say it is, and I’m sure that after reviewing the following evidence, you will agree with me!
In fact, the author of the book on symbiosis (ironically cited by the Wikipedia article as evidence that the broad definition is “widely accepted by biologists,” even though it actually says the opposite!) actually says: while “many symbiosis researchers use this [broader] definition”, “the [broader] definition is not accepted by most general biologists and nonbiologists today, and so fails to communicate effectively,” and “the definition of symbiosis widely accepted among both general biologists and the lexicographers who prepare English dictionaries is an association between different species from which all participating organisms benefit.” (Emphasis author’s.)
As cited by the Wikipedia article:
> The definition of symbiosis was a matter of debate for 130 years.
> [...]
> In the 21st century, the latter has become the definition widely accepted by biologists.[13]
[13] Douglas 2010, pp 5-12:
> This brings me to the most frustrating difficulty in the field of symbiosis - the lack of a single universally accepted definition. Disagreement [...] has led to [...] a lack of consensus. Two alternative definitions of symbiosis, neither fully satisfactory, have dominated the literature for many decades: "symbiosis as any association" and "symbiosis as a persistent mutualism."
> [...]
> 1.3 DEFINITIONS OF SYMBIOSIS
> 1.3.1 Symbiosis As Any Association
> The term symbiosis was coined originally by Anton de Bary in 1879 to mean any association between different species, with the implication that the organisms are in persistent contact but that the relationship need not be advantageous to all the participants. De Bary explicitly included pathogenic and parasitic associations as examples of symbioses. Many symbiosis researchers use this definition and, without doubt, some colleagues steeped in the symbiosis literature will have objected to the opening two sentences of this book.
> One key advantage of the definition of de Bary is that it promotes a broad context for research into symbioses. It acts as a reminder that it is important to investigate both the costs and the benefits to an organism of entering into a symbiosis (see figure 1-1a); and it is reasonable to expect some of the processes underlying relationships that are classified as mutualistic and antagonistic to be similar. For example, just as the persistence of certain antagonistic interactions depends on one organism failing to recognize the antagonist as a foreign organism, so some organisms may be accepted into symbioses because they fail to trigger the defense systems of their partner and not because they are positively recognized as mutualists.
Nevertheless, the definition of de Bary has two serious shortcomings. First and very importantly, the definition is not accepted by most general biologists or nonbiologists today, and so fails to communicate effectively. Most people do not describe the current malaria pandemic or the potato blight that caused the Irish famine of the 1840s as examples of symbiosis. Second, there are few principles generally applicable to symbioses, as defined by de Bary, but inapplicable to other biological systems. As a result, the "symbiosis as any association" definition is something of a catch-all category. Although this definition does promote further enquiry and insight into symbioses, any insights obtained are unlikely to be common to all symbioses defined in this way.
>
> 1.3.2 Symbiosis As a Persistent Mutualism
> The definition of symbiosis widely accepted among both general biologists and the lexicographers who prepare English dictionaries is an association between different species from which all participating organisms benefit. I subscribe to this definition even though it is not without difficulties.
If anything, I would think perhaps that we might be able to agree that the Wikipedia article you referenced should be fixed to correctly represent the author’s conclusion, or the offending section excised. :)
What is striking about written documents from that era is the lucid and yet far more sophisticated writing style compared to contemporary publications.
Would Licklider be happy with the AI slop that we see today? Would he be able to use LLMs in a productive manner?
People in that era could be optimistic because they did not suffer from information overload. The best symbiotic relationship is still man-book or man-pencil-and-paper.
They could be optimistic because they didn’t get suckered by what tech has become. It is the same kind of optimism that can be found in Engelbart’s preamble to the mother of all demos.
The idea that one could treat a program delivered by google, facebook or amazon as something to cooperate with rather than use warily is much harder now than it was then.
Alan Kay chats / blogs / whatever-it-is-people-do-there a lot on Quora and the question "What would be the place of AI in Doug Engelbart's vision of 'Augmenting Human Intellect'? How can we reconsider things within current AI 'developments'?" recently came up.
Thanks for the pointer, didn't know that Quora had threaded discussions.
Excerpt from Alan Kay's answer:
One big distinction in the 1962 time period is that they thought of “machine intelligence” as being a kind of complimentary set of thinking tools that could be “symbiotic” to how humans were able to think. They were not at all thinking about something like a slave or a major domo, but something more like a research assistant or a “Memex” (the latter was a big influence on Doug’s thinking).
In the very late 60s the “official AI researchers” started to think that something like “intelligent Greek slaves” were needed for the “Romans” (Americans), and became rivals to Doug’s notion of elevating human thinking rather than just elevating power. This was a bad idea then … and it’s a bad idea today.
It was interesting to see that Christopher Nolan featured Vannevar Bush prominently in the recent Oppenheimer film. Stripe press published some of his writings recently as well. (I haven't read this one yet)
Major fields of study like chaos (sensitive dependence on initial conditions within an apparently deterministic system), protein structure solving from X-ray diffraction patterns and many others would be impossible without computational assistance from machines.
As far as style and tone, you can certainly get LLMs to mimic Licklider's style with suitable prompting, e.g.
> "In replicating the reflective and anticipatory tone of the original text, it becomes evident that the concept of man-computer symbiosis was not merely speculative but a forecast grounded in the trajectory of technological and computational advancements. This partnership, as envisioned, holds the promise of amplifying human potential through the strategic leverage of computational power, thereby redefining the boundaries of what can be achieved through the confluence of human and machine capabilities. The profound foresight in recognizing and articulating such a partnership underscores a remarkable understanding of the complementary strengths of man and machine, heralding a new era in the evolution of technological augmentation."
That is a pretty grounded, timeless write-up for 1960.
5.4 The Language Problem
The basic dissimilarity between human languages and computer languages may be the most serious obstacle to true symbiosis.
LLMs unlock the much sought after capacity to query data in natural language, that is why "AI" is such a hype, besides not being intelligent (in terms of reasoning) at all.
Licklider was primarily referring to the need to suppercede punch card languages with ones with an interactive human language interface. Chiefly, McCarthy's Lisp, which he directed the Arpa/IPTOs funding towards, but it would apply to practically every language since - though in terms of interactivity still CL most of all.
That said he also lead the funding of AI, so LLMs are not outside of what he was imagining altogether, for example he wrote about the need for a busy executive to communicate with his computer like he would his secretary, etc. But he also seemed to think that interactive programming environments could accomplish this goal.
Sidenote, something that's forgotten is that Lick was a Lisp person. His archives contain his notes to SICP, and various toy programmes.
They don't really have to. Even in Star Trek where you can tell the holodeck to do whatever in LLM fashion, there's also a lower level language that can be used for more fine control. It's always good to have levels of abstraction when you need them, this just adds the very top one that was always missing.
The fact is that they do and it doesn't matter how many abstractions we put on top of the underlying physical reality of digital computation, it will always be digital unless someone figures out how to make analog computation as scalable as digital computation.
I assume that the reason this is relevant now is because of the the announcement of FinalSparks NeuroPlatform, that gives SaaS access to brain organoids?
Very cool but also very creepy. Has a clear potential to become the most evil thing humanity has ever done.
Checking now, it has been posted to HN a few times in the last weeks but without discussion really taking off.
* https://www.goodreads.com/book/show/722412
reply