Worse, still, unless the current me blinks out of existence, "old me" will still be around - unless we can replace neurons in-place with computational analogs, the original me is destined to die, almost no matter what, even if a copy/download is made. From the download's perspective, it continues, but the original, if not ended then and there, still faces a grim end and will know it.
How is that all that damaging? I’m sure those batteries are recycled in basically all cases and yes, I’d expect battery replacement does indeed forestall replacement for many phones.
The most common processes for recycling lithium ion and lithium polymer batteries are not environmentally friendly. They consume a lot of water and energy and produce toxic byproducts the require further processing and energy to be rendered safe.
I think embodiment and the encoding of the natural laws (gravity, force, etc) that go into that will be another huge step at grounding AI. People tend to gravitate to thinking about humanoid robots when that is mentioned (and thereby terminators), but honestly I would think things closer to sensor networks involving thousands or millions of bodies like a hivemind would be more likely (why stick at the human level of a single body if you didn't have to). Interaction with the world is a means of determining truth... The ability to perform science.
And as hard as embodiment is, it will be the easy part in my opinion. Continuous learning without losing the plot is going to be quite the challenge. If an LLM has something wrong, how does it update and change that bit of information without huge amounts of power use? How do you make the system learn 'important' things without filling up with junk/spam it is subject to? How do you keep the system aligned with a goal that is not destructive to itself or others?
But embodiment being a bottleneck could indicate that it's a data/training issue, rather than an architectural issue. Multimodal training data improves GPT-4 already, but that's still very little data compared to growing up to a full human adult. There are still many things to try.
That has always been my impression, despite the myriad ways that LLMs impress.
So much potential is lost just in the request/response limitation. While I’m waiting for a response from GPT-4, I’m continuing to think. Imagine if the reverse were true. AGI needs to be able to mull things over for spans of time.
At least any company trying to sell a product this is going to be an issue with operations costs.
Also this gets into the halting problem. How many resources do you expend on finding an answer? In a human issues will typically come up like we have to go pee, or eat, or something outside our body interrupts us. For an AI, how much time should it spend? Do we want to wake up one day finding our data centers running at full tilt?
This said, there have been some attempts at working on agent based systems that reach out for answers from multiple places and pool the data then run things like chain of thought on that data pool.
Enshittification is such a cruel thing. Arstechnica was a quality site at one point… not to say it has completely lost its value, but it’s sad to see the its decline, nonetheless.
We have familial ALS in our family. I lost my grandfather, two aunts, and uncle and a number of more distantly related family members to this disease.
Thank you for posting this article – I'm lucky - my dad tested negative for the gene that has afflicted our loved ones – but I feel so guilty vs. all of my cousins who still have to face these risks.
We have the power to enable them and others like them to make their own decisions, to have even a slim bit of hope (or even a slim bit of uncertainty) for the first time since we've recognized that familial ALS was a thing. I can't overstate how big of a deal that is. Agency, even if risky, is so much better than the fear and powerlessness of waiting for the inevitable.
The problem with most self checkouts is that the value prop for the stores seem obvious - lower wages - but the value for consumer is not obvious in most cases - because in most cases, it doesn’t exist.
It’s not a better experience and it feels very much like I am now being asked to do more while still paying more.
That’s far from a unique value prop for self-checkout lanes. It’s also a hard sell at stores that have or had ~8+ lanes but only two are open… sure, the self checkout is the only solution to long lines… that happens frequently here and in many cases, lines get equally long for self checkout.
It’s so severe that even if they improved the terms, nobody would use them because US debit cards' reputation is tainted.
Say what now?
“According to the 2019 Federal Reserve Payments Study1, “Debit cards, including both prepaid and non-prepaid, were used almost twice as often as credit cards in 2018, but the value of credit card payments exceeded the value of debit card payments by almost 30%.”
And
In Visa’s Operational Performance Data4 report for the three months ending on Dec. 31, 2020, it reports $741 billion in U.S. debit transaction volume, a 17.4% increase over the previous year. It compares to $542 billion in credit card volume.
They used to sell the case, and each airpod separately (with the total for all three being the same as a new pair), but it looks like they stopped? It was very useful when I ran over my airpods case with my LandCruiser, and it got a bit wonky (but was amazingly still fully functional).
And I wish I could just trade the Type-C case with my lightning case (with at most a tiny refurbishing fee), there are a lot of people who just bought the Airpods pro 2 last or this year and not likely to purchase again in 2 yrs
