gwright

If we are going to expand the discussion to general inaccuracy in human communication we are going to need a much larger thread.

I do agree with you that "serverless" is a misnomer and not a useful technical term. I'm guessing the marketing team wasn't fond of "dynamically provisioned and auto-scaled pool of virtual machine instance-based services".

I think "fabricate" is much less aligned with cognition.

It is completely reasonable to say that "This machine fabricates widgets." or "This algorithm fabricates pseudo-random song lyrics." without being worried that the machine or the algorithm is exercising "cognition".

Maybe you are a large language model?

That is an interesting point. I'm not sure why, but "hallucinate" doesn't bother me as much as "lie". Maybe it is because of all the ancillary baggage that "lie" has related to our current culture wars (lies, misinformation, disinformation, gaslighting, etc.)

I just asked (another anthropomorphic verb) ChatGPT for a better word and it came up with "fabricate".

> The term "fabricate" could be used to describe AI's tendency to generate fictitious information. This term has less anthropomorphic connotations than "hallucinate" and directly refers to the creation or construction of something, which could be a suitable representation for what an AI system does when it generates false data. Other alternatives could be "synthesize" or "generate," depending on the specific context.

I think it confuses things immensely to anthropomorphize large language models. LLMs don't lie or tell the truth they just spit out text that is in alignment with the training model. Don't give them agency they don't have.

I noticed something similar while visiting Texas last year. We have an Amber alert in my home state but it is triggered maybe 3 times/year.

When I was in Texas I received the alert 3 or 4 times in a couple of days. The people I was visiting seemed to think that was typical.

It seemed to me that the alerts weren't very localized.

I'm a little rusty in this area, but I'm pretty sure you'll have a hard time getting a direct IP allocation (vs from your transit provider) unless you are multi-homed, which apparently they were not.

There is a nice tick up in complexity when you go to advertising your address space via BGP to multiple providers.

> Its a language model and it fooled DH. It hasnt gotten smarter than us yet. Its just faster at repeating what other humans verbally said.

One of my take aways from LLMs is that humans very often just repeating what other humans have said with only superficial understanding of what they are repeating.

I think there is more to general intelligence than pattern matching and mimicking but a disconcerting amount of our day to day human interactions might just be that.

Hofstadter makes the claim that "these LLMs and other systems like them are all feed-forward". That doesn't sound right to me, but I'm only a casual observer of LLM tech. Is his assertion accurate? FWIW, ChatGPT doesn't think so. :-)

FWIW, we started using https://conventionalcomments.org a while back and found it to be a useful tool for making it easy to communicate what is an issue that needs to be fixed, a suggestion to consider, or a nitpick to optionally ignore.

It is still just a tool and does require the team to buy-in and utilize it, but we've found it helpful.

Nope, I'm actually pretty cynical about intentions. But the government can detain me, arrest me, tax me, fine me, etc. A corporation can't do any of that.

I'm not sure what you really mean by "monopoly" or what sort of monopoly you think can persist in the absence of government support.

Note, I started my participation in this thread by advocating for minimizing government, which is not the same as eliminating government. So I'm not trying to assert an absolutist position that free markets should have no regulation.

Ok, now explain how free market economies have improved the quality of life and lifted more people out of poverty than any other economic system in history.

Monopolies can only exist with government support via protection from competition. Expansive government power makes monopolies more likely and more stable than minimal government power.

How do you define ‘power’. The unique aspect of government controls is that we grant the government a monopoly on force (imprisonment, seizure of property, etc). That is not the case with private actors and that is a significant and consequential difference in ‘power’.

One reason to advocate for a government with minimal power is to avoid the temptation to legislate preferential treatment. If the government has been granted more expansive powers, it is inevitable, and arguably rational, for those affected to try to steer regulations in a way that benefits them.

> In America, the socioeconomic class of people of non-white races has been systematically kept as low as possible, through racist systems. Affirmative action (race based admissions quotas) is one tool to breach that ceiling.

According to this: https://en.wikipedia.org/wiki/List_of_ethnic_groups_in_the_U...

The per capita income for "whites" in the US is $36,962

The per capita income for the following groups is higher than that: Indian, Taiwanese, Japanese, Chinese, Korean.

Similar results for median household incomes.

I couldn't easily find the data but I've seen previous reports of sub-groups of "blacks" that also have higher median incomes than whites (as a group). Ahh, here is an article talking about the success of Nigerian immigrants, as an example. https://journals.sagepub.com/doi/full/10.1177/23780231211001...

Please explain how to reconcile this data with your statement.

> Burn it all down imo and let something else take its place.

Probably hyperbole, but catastrophic failure of our economy/institutions/society isn't something I would choose to experience.

This nihilistic attitude is dangerous, IMHO. In the extreme, it is a self-fulfilling approach with severe consequences.

Seems like we should be able to do better than that as a society.

Right, because nationalized industries/companies have a wonderful track record of environmental concern and practices?

I'm not sure what the solution is to these problems (or this particular problem) but "nationalizing" producers certainly isn't one of them. Destroying 3M isn't one either.

I don't understand the approach to difficult problems that starts with thinking "the government" is effectively a magic wand.

I don't know how to easily quantify these things but your analysis ignores the other part of the transaction. The consumer of the photos (or ML generated text) benefits through the lower cost, easier access, increased project velocity, etc.

Market disruption due to technology/innovation/progress is nothing new. There have always been winners and losers but the aggregate wealth/value continues to climb. Devising ways to mitigate the loses seems like a more feasible approach than pretending that we can prevent these sorts of changes through regulation or legislation.

This makes no sense to me. Tech workers have way more flexibility than many other types of workers. If you aren't comfortable with your employers practices, go find another employer. No need for a union to do that.

> Mr. Geist also accused the government of burying the change, by putting it near the bottom of a nearly 450-page budget bill last spring. The government didn’t highlight the copyright act changes in any of its documents about that bill.

So it looks like it was "debated" in an assembly. I'm using scare quotes because I'm not a big fan of burying potentially controversial changes in giant budget bills. I'm also not a fan of giant budget bills or giant bills at all for that matter.

All things are "the same" at some level of abstraction. But if you have to reach to a high level of abstraction to illustrate that the things are the same, I don't see that as a particularly insightful comparison. Sameness is a mundane observation at a high level of abstraction.

Sort of like just tossing your hands up and saying "life sucks" regardless of what "system" you are living under and then concluding all "systems" are "two sides of the same coin".

That is quite the stretch to think those situations are "a bit [a]like".

That math doesn't work -- at least until someone can demonstrate cost effective grid scale energy storage systems. In other words it is not cheaper and perhaps not greener (it isn't clear how green the production and long term maintenance of solar/wind is)

The basic problem is that in order to provide reliable energy 365 days a year you need to account for the daily and \ seasonal variability of solar and wind. You have two options:

    * maintain a duplicate fossil fuel backup system (fully maintained, staffed, and fueled) for those cold windless nights (so no longer cheaper)
    * deploy a storage system that can effectively time shift energy created by solar and wind across entire seasons (also not cheaper nor demonstrated with current battery technology)

Of course if you don't care about reliable electricity you have one more option:

    * suck it up when there is no sun or wind available

Thank you for that example. I've been thinking a lot about how difficult it is to persuade people to change their mind. Most people seem impervious to incorporating new facts or ideas into their internal narrative (myself included). ChatGPT has made me wonder if that is because much of what we consider "cognition" in humans is really just "human sentence predictors".

We've all been "trained" with various facts and when two people meet who have been trained on substantially different bodies of knowledge/facts/experiences it can be very difficult to find common ground.

FWIW, I asked ChatGPT to give me a short list of cognitive biases and psychological phenomena. Interesting to think how many of these are dependent on our personal "training data":

    Confirmation Bias
    Cognitive Dissonance
    Anchoring Bias
    Belief Perseverance
    Groupthink
    Ingroup Bias
    Sunk Cost Fallacy
    Motivated Reasoning

I feel like your observation that this "isn't a complicated question" is leaning on an implicit assumption that ChatGPT is a general AI and not a LLM. It is just generating text based on probabilities -- it isn't "reasoning". I might go as far to say that inferences computed by a LLM are of all the same complexity but I don't really know enough about ChatGPT to be confident in that statement.

> But the internet seems a very biased sample.

It also seems to me (acknowledging my lack of expertise) that LLMs trained from online resources are likely to weight text that is frequent vs text that represents "truth". Or perhaps I should say repetition should not be considered evidence of truth. I have no idea how to drive LLM models or other ML models to incorporate truth -- humans have a hard time agreeing on this and ML researchers providing guided reinforcement learning don't have any special ability to discern truth.

> I disagree with your confident assertions about its "agency" and "intent" when there's no goal post for what those things mean in humans to begin with.

I'm not sure how confident I am!

I think the success of LLMs are serving to challenge our understandings of cognition and intelligence in humans. I tend to agree with you that our understanding of agency, intent, intelligence, cognition, free will, and related ideas in humans is incomplete and so it is a challenge to think about how they might apply (or not) to LLMs.

> whats the difference between faking it and actually having agency?

Good question. The Chinese Room Argument is very close to this but framed as "understanding" vs "agency": https://plato.stanford.edu/entries/chinese-room/

> This is not remotely like what human brains do. Your ideas cohere from network connections between the neurons in your brain, and then you come up with words to match your idea, not your previous words or the frequency that the words appear in your brain.

I'm pretty confident that that isn't all the human brain does, but we certainly do that in many situations. Lots of daily conversation seems scripted to me. Join a Zoom call early on a Monday morning:

  Person 1: Good Morning!
  Person 2: Good Morning!
  Person 3: Did anyone do anything interesting this weekend?
  Person 1: Nah, just the usual chores around the house.
  etc.

All sorts of daily interactions follow scripts. Start and end of a phone call, random greetings or acknowledgements on the street, interactions with a cashier at a store. Warm up questions during an job interview...