Inevitably the answer to any article title comprised of a question is "no".

One day perhaps, if the distant future if/when we have truly animal like AI with emotions and feelings for others, capable of learning, etc, but for predict-next-word functions? Perhaps we should assign personhood to sort functions too?

Just because an LLM outputs human-sounding text (well duh, its a next-word predictor) doesn't make it any more like a person than the cat command when you do "cat mythoughts.txt".

Yeah, even with an EV with longer range I'd get range anxiety if relying on finding a charger along a longer trip... Until charging becomes quicker, and chargers more ubiquitous, it seems the best case for EVs is something that range-wise is good enough for your daily usage and can then be charged overnight at home.

Too bad the base model is a pickup vs "SUV" with second row of seats, but it seems a pretty decent spec for the price point for a US model (although I'd prefer a $25K BYD if it was available here, and without tariffs that double the price).

Google seems to have deployed a quick fix to avoid these, but Gemini is still splainin.

That's a great expression! "A lame duck still quacks" means that someone who is in a weakened or powerless position, especially because they are leaving office or losing influence, can still make their opinions or demands known.

Think of a lame duck â" it's injured and can't move around easily, but it can still make noise. Similarly, even if someone's authority is diminished, they can still speak up and be heard.

Here's a breakdown of the meaning:

        Lame duck: This refers to someone whose power or effectiveness is limited because they are in their final period of office after a successor has been elected or appointed. They are seen as having less influence because their time is ending.
        Still quacks: This part emphasizes that despite their weakened position, they can still voice their opinions, make demands, or exert whatever remaining influence they have.

So, the expression highlights the fact that even those who are seemingly on their way out or have lost significant power are not necessarily silent or completely ineffective. Their voice can still carry some weight.

All intelligence is prediction. Crystalized intelligence is essentially one-step prediction - you've seen/done if before and remembered, so you can do it again. Fluid intelligence, aka reasoning, is multi-step prediction - chained what-if "tree search" (if one chain/branch of reasoning, then abandon that and try another).

The key to both types of reasoning is pattern recognition - the more patterns you've experienced (= more experience), the more capable you will be. Reasoning also depends on reasoning/logic patterns (what next step does this problem suggest), which is again something that benefits from experience.

So, no, you're not born with maximum fluid intelligence - you can gain fluid intelligence (reasoning capability) by practice and experience if you choose to.

The nature of your comment makes it clear you are young, and never saw it in 1977. Presumably weren't even born in 1977.

Your handle isn't fooling anyone.

It's not just genre studies and similar fare - degrees that pay for themselves, such as law and medicine, are the exception. Even something like Computer Science, while it might help land a job (at least until recently) isn't actually teaching you much of value - you'd be better off leaning on the job (as IBM Used to do - preferring to teach programmers the IBM way vs having to unlearn bad habits they may have picked up elsewhere).

We really need to get away from expensive degrees as a form of gatekeeping, and the expected norm, and go back to being people receiving on-the-job training as apprentices and juniors.

It'd get the debt off your back, but it's not going to get you a job.

There is no connection between AlexNet and Transformers, so I'm not sure why you are considering them together. They represent totally different kinds of breakthough.

AlexNet's historical importance is in kick-starting the modern "deep learning" neural network revolution. It wasn't an architectural breakthough, but rather a demonstration of what could be achieved at scale and using GPU acceleration.

The interest in the AlexNet source code is presumably because that is what was special about it - it was a hand coded CNN, pre-cuDNN, demonstrating that a large scale CNN was feasible and could be trained on a COTS GPU. It's impact on the computer vision field was in ending the line of research of hand designed vision features like SIFT, and letting feature representations instead be learnt.

The significance of the Transformer is as an architectural breakthrough supporting the efficient training of extremely large and performant language models, which it turns out can do a lot more that one might expect of a language model!

Apples and oranges.

There are of course many small models whose weights you can download and run locally, but typically training data isn't provided (nor training instructions), so whether you consider that "open source" is up to you. Of course most people don't want to (or can't afford to) incur the cost and effort of training a model themselves, so weights is all they really want.

Now, to be useful an LLM needs to be large, so training it ONLY on your personal data is really a non-starter, unless you somehow have a self-authored equivalent to WikiPedia, Reddit, Stack Overflow, etc, etc.

But, maybe you're more concerned about your own personal privacy and data security, and don't care about using other people's data, in which case just go ahead and download a local LLama model or whatever.

Only if that's what you want.

No reason they couldn't support jive too if that's your preference.

https://fahdjivetranslator.net...

That's true of Tokamak type designs, but Helion's approach seems much less problematic.

I'm not sure that rockets are ever going to become much safer than they are today.

I'd like to got to space for sure, and doubt it'll ever be affordable for a normal person in my lifetime. I'd probably be willing to pay $100K max to go. A free trip would be nice.

Where and when is another question though. Given the risk, the reward has to be worthwhile, and Blue Origin doesn't really seem worth it - could have a much safer MiG flight to the edge of space and see a similar view. A week on the ISS would be worth it.

I've got a daughter yet to enter college that I'd like to see on her own two feet before I kick the bucket, so probably wouldn't accept the free ride right now, but at the right time I think I'd do it. Maybe Mr. Beast would send someone.

Sure, "AI gone rogue" is a money-making headline, but there is some truth to the danger of RL-trained paperclip-maximizing AI.

If you really want to drive customers away from talking to a live customer support person, then why stop at a 15 min wait? An hour or so would seem better, and would have the added benefit or tying up the support lines so other callers just get busy tone.

But, why stop there? Some people may be ok with being put on hold for an hour, and maybe even enjoy the elevator musak you play to them.

To really drive the callers away, it'd seem better to make it a truly horrible experience. Instead of elevator musak, how about a series of load jarring sounds - explosions and people screaming - played at eardrum busting decibel levels, and interspersed with someone screaming things like "go kill yourself", "get the fuck off the line", etc?

It's a tool. It's as useful as you make it.

You can use a power drill to drill holes in the wall, but also to drill holes in your foot, or try to use it as a hammer. The person who chose to use it to drill holes in the wall is more likely to regard it as useful than the person trying to use it as a hammer.

Stating something as your personal opinion does not magically elevate it to an objective fact. The things that people have accomplished with AI speak for themselves - those are objective facts.

If you, personally, have not been able to get AI to do what you want, then that's just a reflection of your choice of attempted tasks and how you're wielding the tool. Those experience are objective facts too.

Where you make zero sense is in asserting that your own failures mean that others must be failing too. That's just arrogance and stupidity. It's like you tried to run a 4 minute mile, came in at 10 minutes, and then confidently stated "Running a 4 minute mile is impossible. That's not an opinion, that's an objective fact."

I guess you'd most likely throw some italics in their too, and resort insults if you got frustrated by someone else pointing out that running a 4 minute mile is, in fact, possible.

I'm amused to see your continued responses. Evidentially your time isn't quite as precious as you think it is.