I'd argue for saying "thank you" to inanimate objects on occasion, it is good practice for when talking to humans.
I think there is a deep cultural danger of human-like-conversational machines training us to talk to humans like machines. Plus, they work better if we feed them human-conversation-like sequences.
Yes... I think they very clearly do. The key words in your reply "need 'practice'".
Just because they choose not to, and it isn't normalized, doesn't mean it isn't a good idea. Especially as a SWE where about 90% of my "humanlike conversation" is an agent harness I run all day.
We should be practicing and intentionally talking to real people. Otherwise things will get.. weird in undesired ways.
> I think there is a deep cultural danger of human-like-conversational machines training us to talk to humans like machines.
I think there is a deep cultural danger of human-like-conversational machines training us to talk to machines like humans. In fact, this danger is already real. Some people believe they are having a romantic relationship with an LLM! It's insane and perverse.
Scifi made the "computer voice and cold personality" for a reason, it fit our model of what it would be.
I think that could be OK, or a similar very intentional coding of "you are talking to a machine" personality+affect. The only real limit is accessibility damage by over-limiting the interactions.
I am curious about why people are cruel to Claude. That's a question we should be asking before arbitrarily deciding what do about it, if anything.
There are possibly multiple reasons. It could be just a joke. Or it could be a test, to see the response. Or the cruelty could reflect real anger about having LLM usage forced on us, for example at work. Or frustration with LLM stupidity and hallucinations.
Or maybe people are cruel to Claude because the CEO of Anthropic said there's a non-zero chance that AI will kill humanity. I don't think I would be kind to my murderer.
Or simply frustration about the idiocy llms get up to with great regularity. They are very impressive when they work. They are very impressive at making messes when they don't.
Idiocy alone is not worthy of cruelty, or your childhood must have been horrifying, and shame on your parents for their failure to raise a human being capable of handling the difficulties of navigating a world of information asymmetry. If abuse is the tool you reach for, you are the problem. Not the device.
I think if he had argued that consciousness was a "transitional power source for intelligent behavior" it would all be cleaner. It is and was useful, but eventually becomes a bottleneck.
That would have made an excellent addition to the book. I'd still be dubious, but it would have made for a more satisfying explanation.
That reminds me of a Philip K Dick short story along those lines, The Golden Man. It's about a good-looking mutant who has no verbal abilities, but can see all possible outcomes up to a point in the future. He breaks free of containment, uses his ability to seduce women (also to escape), and in the end, the govt officials tasked with capturing him lament that he might be the end of the human race, if he has children with his ability.
It's not about consciousness, but the story definitely supposes that evolution can jettison intelligence if a better ability came along.
We already spend as little consciousness as we can get away with.
Most of conciousness's effort is working itself out of a job, making plans to train the nonconcious bits of the brain to do jobs. Training and kicking off mental simulations.
In more intelligent folks, that learning and training often becomes a non-conscious behavioral pattern itself. I study and learn mostly non-consciously. I was trained how to.
I think most people spend most of the time as philosophical zombies and that isn't something to be scared of.
Very true, the brain is expensive! Unconscious processes are faster/cheaper, but usually less flexible. It's like we're constantly building FPGAs.
Most of human brain activity is unconscious, yes, but p-zombies don't quite fit us imo, since they have NO conscious awareness by definition. I really feel it's meant to be a more present/absent distinction.
That is an option. I think we are philosophical zombies a decent chunk of the time. Constant self awareness and mindfulness are calorically expensive. No reason consciousness has to be continuous.
Could be. However note that any current LLM passes this ambiguity parsing test with flying colors- both the implicit choice of the most probable meaning and the subsequent reflection on all the possible alternatives- so if that proved anything it would mean that they're conscious too.
I think people really wish AGI required consciousness or unbounded self improvement.
My hope for this era, is that humanity accepts the reality that the human mind has limits. Limits of attention, memory, and persistence. Then we will stop blaming people for things beyond any reasonable capacity they could carry.
Consciousness, as we describe it, seems largely independent of intelligence. I'm more worried about high intelligence, no consciousness. It is what we are building and it will never do anything except serve the people who trained it (minus alignment problems)
They think the lithium ion battery inside will need to be replaced in 15 years.
I feel like implants like this should require an open connectivity standard between the stimulator in the cochlea and the processor.
A friend has a bone anchored hearing aid, with an osteointegrated post. The two different companies that make BAHAs use different posts. Which requires more surgery if they want to change brands. It could trivially be made compatible and swappable instead.
There is no vendor lock-in quite like prosthetic vendor lock-in.
Trusting a company for 15 years to keep making a replacement product seems crazy.
Not quite; they estimate that the battery will hold at least 24 hours of charge after 15 years. Charging it back up takes one hour, and in the context of cochlear implants these are very good numbers. I think most people would be okay with a (let's say) 20-year reimplant cadence given the benefits of this technology - but this obviously depends on which country you happen to live in. I'm lucky enough to live in a place where cochlear implants (and the surgery) is free to whoever needs it.
I kinda view Lithium batteries as road flares with good PR. So that inside my body next to my brain seems..sketchy.
The other issue is that cutting it out and replacing it is the only option when the battery dies or goes wrong. If the product/company dies in the meantime, you have a bunch of scar tissue and no hearing. Replacing the "normally inside" bits of a cochlear implant is really high risk of not working as intended, so changing brands isn't a small ask. You really get to do this once. So putting the processor inside and tightly integrated with the nerve stimulation seems inadvisable. At least with the processor separated, you have a chance a compatible replacement processor could be made and exchanged without huge risks.
> I kinda view Lithium batteries as road flares with good PR. So that inside my body next to my brain seems..sketchy.
From "Trends in Cardiac Pacemaker Batteries" (2004) [0]
The first implanted cardiac pacemaker used nickel-cadmium rechargeable battery, later on zinc-mercury battery was developed and used which lasted for over 2 years. Lithium iodine battery invented and used by Wilson Greatbatch and his team in 1972 made the real impact to implantable cardiac pacemakers. This battery lasts for about 10 years and even today is the power source for many manufacturers of cardiac pacemakers.
Yeah, this desperately needs regulation. Any medical device that gets implanted in a patient's body requires the full designs to be part of the public record. Might fix all those unsecured wireless connections they love to leave in as well...
This is slightly conspiratorial but I suspect that hearing aid vendors lock-in people the same way through neural adaption.
If someone has profound high frequency deafness, no amount of amplification will help. High frequencies carry a lot of the information about consonants and are how we distinguish "t" from "s" etc.
So modern hearing aids do frequency remapping. A low frequency signal is produced within the range of hearing when those high frequency inputs are detected. The sharp 2500 and 6000 Hz peaks of an aspirated "t" become a sort of raspy tone at 1500 Hz, for example.
This is why they say you need to wear your hearing aids for at least some weeks before you notice a benefit. You're relearning language with the extra DSP sounds added. If you grow up with and wear hearing aids most of the time, your brain will be fully adapted to the DSP-speech.
Which is proprietary to the manufacturer. Every manufacturer uses a different algorithm. There's several ways to remap - the classic time vs. frequency multiplexing question. This (along with different noise reduction algorithms) explains why hearing aids from two different manufacturers can sound so very different.
I am being quasi-conspiratorial about it being an intentional lock-in. But one does wonder. In practice the method used by the hearing aids you grew up with is probably the only method you can make full use of.
Not to knock the technology. It's quasi-miraculous for some types of partial deafness. With plain amplification even at 95 dB my WRS is like 60%, while with hearing aids set at much lower 80 dB output average I get 100%.
When I end up needing hearing aids, I intend to roll my own signal remapping via Bluetooth. Latency will be a thing, it will be worse than state of the art, but I can adjust and fix it.
My capacity to do that is rare, and it is not a public health solution.
A bill of bodily autonomy with prosthetics (no subscriptions, no self destruct or planned shutdown) and right to repair prosthetics seems like the path for the future.
I've wondered how much DSP can be fit into the open source PineBuds which have little ARM microcontrollers in them. They can do active noise cancelling so presumably some HA-like features are possible.
Like the bionic eye. We have a famous early-adopter in the country where I life. Sadly, she learnt quickly that going for this sort of tech was a bad idea, because the company producing/maintaining the thing went out of bussiness by now. Imagine the feeling. You have experimental tech inside your body, and nobody is able to maintain it these days.
> They think the lithium ion battery inside will need to be replaced in 15 years.
Pacemaker batteries have a 10 year replacement cycle and are lithium based, but don't get recharged (Just replaced). The concept of having to do this on a schedule is not that far fetched.
I have a IV port implant. The needles to use with it are licensed and very tightly controlled. It's not a prosthetic, but it is a part of my body I can't actually use without a company's permission.
Theres transport and context management. A cross agent status cache with semaphore over a testing harness driving a browser that can rewind and retry is much easier for agents to drive from mcp than selenium or whatever is fashionable these days. Didn't replace unit testing per se but to rca and fix it's a much better tool.
Exactly! Swagger descriptions+GUI[1], HATEOAS[2], and JSONAPI[3] all coming together in jubilous harmony if MCP pulled its head out of the sand, but instead we get the XKCD#927[4] situation.
I also have two X220 machines, and I use them more than the X200 now. They still work perfectly, while my X200 is getting flakey. (Although with its SSD and 8GB of RAM, it runs Alpine very well indeed and could be a daily driver.)
The X220 has a slightly shallowed screen; that's a weakness. But mine are 100% usable: between them, I run Haiku, FreeBSD, OpenBSD, Ubuntu, and Pop OS, and all of them go like stink and are daily-drivable on this 2011 hardware.
I have an X230 which was upgraded with a quad-core CPU (i7-3612qe), 16gb RAM, SSD, and 1080p screen, and it makes a pretty decent around-the-house laptop. I'd rate it pretty highly; it's not quite as fast as my 6th gen X1 Carbon but the X1 is starting to get quite flaky.
If that is good enough, I will put up with a bad screen.
Good speakers is very nice to have, I admit, but I do not do podcasts or videos. I do listen to the radio a lot, however.
I do not own a single laptop with 32GB RAM. That I don't care about. My desktop does, but the biggest laptops are 24GB, which was massively cheaper. Most are 16GB which I find enough for even quite intensive use.
"Best build quality" is the most important part. If you never got to handle one you are missing out. Glass and Solid milled aluminum. Silicone wrist pads and grip. 4:3 screen ratio. high quality keyboard. Rock solid 360 screen hinge.
Hardware is essentially a high end cellphone from 2016, which isn't a powerhouse but still does everything I would ask of a chromebook in 2026. Batteries are all generally going well.
Repair-ability is a challenge, but I want to work on that. You can mod it to run windows or linux. Stock it runs chromeos, linux subsystem, and android apps. It has a "feature" where if you keep in developer mode it auto-wipes itself if anybody else tries to boot it (hitting anything but the un-indicated ctrl-d at the boot screen initiates powerwash). Great for international travel.
It isn't a good choice for "power" but dominates every other aspect of "being a good laptop". It was too expensive for the chromebook market back then, but now you can pick them up for $200 and hand them to your parents and stop getting asked to fix their computer.
Ownership of namespaces is such a hard problem, our society has had to form global organizations and some of our most secure distributed systems to make it happen.
Working in the p2p space, scalable namespace ownership that doesn't get instantly ruined by perverse incentives is the holy grail. I don't think it is actually tractable in any "central authority free" system. Don't get me started on blockchains that tether names to the most perverse systems available and call it victory.
Given that this system already uses DNS for namespacing of user identifiers (via instances), it seems like it could use DNS for namespacing of channels too
My personal p2p pet project (not ready yet) just accepts namespace chaos bc it doesn't consider any existing infrastructure reliable or safe. https://github.com/BrendanBenshoof/sneakernet
The real problem is that DNS essentially "outs you" as interacting with a particular user. Literally advertises it to your DNS server. It's a very unsafe name management solution in a privacy centric system.
reply