It's certainly capable of it.

https://matemarschalko.medium....

I'm sure there's folk who would pay enough to make it a viable product for OpenAI/Microsoft to release in this form. Would be useful to have it on phone as replacement for Siri and Google assistant too.

Subway sandwiches aren't that bad. No great, but certainly not disgusting, unlike the McDonands burger I foolishly chose to have for lunch yesterday.

Certainly an interesting pivot or new direction though. From sandwiches to EV charging!!

It does make a little sense to extent that EV charge time is about as long as it takes to scarf down a subway sandwich, but got to wonder how may Subway locations have room for this, or could even justify the cost of a single EV charge point in the parking lot (if they have one). They tend to be located in high-street type locations where there's no space.

Everyone is talking about AI, so he want's to insert himself at the center of it all.

AI will become a potential risk only when anyone let's it control anything, as everyone is well aware. No need to sound the alarm over a chat bot or code auto-complete tool.

Maybe Musk should worry more about his Teslas not killing people, and leave AI up to people who understand it.

What's "dumb as bread" are users talking to ChatGPT who don't realize what it is and have unrealistic expectations. If you treat it as an "AI" or all-knowing oracle, you will be disappointed. It is just a language model.

It is useful if you play to it's strength as something that knows about language.

They're not trying to control what the user can ask ... They're trying to control how the model responds.

ChatGPT, despite all the "AI" hype is fundamentally nothing more than a (very powerful) language model, or what you might think of as an auto-complete engine. You feed it an initial "prompt" (any text - doesn't have to be a question), and it will reply with the sequence of words it calculates would be statistically most likely to follow your prompt.

Each user input and ChatGPT response (i.e. the entire conversation) become part of prompt that subsequent responses are based on.

What Microsoft/OpenAI are *trying* to do with these initial inputs (which become the start of the conversation) is to nudge ChatGPT in the direction of generating output that is consistent with (i.e. statistically likely to follow) these initial "instructions".

There are lots of ways an LLM like ChatGPT could be combined with search to improve the experience, and the interesting question is what are Microsoft/Bing doing here?

I think most people might be assuming that ChatGPT is being used to generate search query "content", the same as when ChatGPT is used stand-alone, but there's a couple reasons that seems unlikely to be the case:

1) Apparently Microsoft's ChatGPT integration is able to cite sources, which is something ChatGPT itself is fundamentally unable to do. The technology behind ChatGPT is what's known in machine learning as a "language model", or in this case a *large* language model (LLM), which is something that essentially deals in word statistics, not facts or sources. ChatGPT just generates words one at a time based on the coimbined statistics of everything it was trained on. As far as ChatGPT is concerned there is no source for the word-by-word output its generating - it's just a product of the meat grinder if word statistics it was trained on. Occasionally the output of ChatGPT might word-for-word match a source, but that is coincidence and not something that AFAIK it is currently designed to detect.

2) People expect search results to be up to date, but ChatGPT is based on a frozen training set maybe a year old (and it costs tens of millions of dollars to retrain, so that is not going to be done often). There is a line of research into having a frozen ChatGPT generate queries to be used as *inputs* to search engines (and other sources), which is conceivably what Microsft is doing if this is capable of returning recent data.

So, if ChatGPT itself, as used by Bing, isn't generating the response content (as such), then what might they be using it for ... There are a number of possibilities, such as:

1) As a chatty front end to Bing's search engine - Bing still does the search, but you interact with it via ChatGPT.

and/or

2) As a summarizer (something LLMs do well) for Bing results - so more than just a front end, but not generating the search query content itself, but perhaps summarizing or rephrasing it in more conversational style.

Given the popularity of ChatGPT it seems we're going to see some type of integration of this type of LLM technology into all the major search engines, and they may all do it in different ways. What they most likely all will have in common is that they won't be replacing a traditional web-crawling search engine with straight-up ChatGPT interaction, since no-one is going to accept something presenting itself as a search engine just making incorrect stuff up (a good portion of the time) and unable to answer questions not based on a frozen training set.

Citing ChapGPT as an "author" is more like admitting to plagiarism - parts of this article copied from sources unknown, or giving TMI about the editing tools you used "grammar corrected by grammar-monkey", or perhaps "I'd like to thank my parents, Clippy, and ChatGPT for the copy-pasta".

The only place I could see for real using ChatGPT in a paper would be for generating a summary.

I have to guess anyone listing ChatGPT as an "author" is just doing it for laughs/attention.

Probably somewhere in the middle in 15-50 year range. Things always move slower than you imagine, even if great progress is being made, and it's depressing to see companies like OpenAI put out parlour trick "AI" chatbots rather than actually work towards intelligence.

Still, neural net advances in last 10 years make me optimistic that we should see real AI in my lifetime.

Unless you're doing embedded, or telecom, or finance, or ...

But, yeah, not going to find much c++ if you're doing windows development.

I don't think the complaint is whether it's well done or not, although there's obviously quite a difference between the typical American QR code to PDF menu experience vs the Chinese all-in-one WeChat menu/order/pay system.

The real complaint is that if you are choosing to get yourself out of the house and away from the computer to the real world full of people, sights and sounds, then many people want a real-world tactile experience where there is a real menu, real waiter/waitress/etc. It's a matter of the restaurant experience, same as the movie theatre experience and all that entails, etc.

For all the right wing whining about being cancelled/censored, it seems they will soon own most of the current (anti-)social platforms.

FaceBook already controlled by Trumpy Zuckerberg, who helped him get elected in the first place.

Twitter about to be controlled by Musk.

Parler controlled by Ye.

Truth Social controlled by Trump.

Pretty much every message board anywhere on any subject overrun by Biden haters who blame him for everything from Trumps orange color to events in 2030 that have yet to occur.

Oddly the centrist/left folk don't really seem to mind (although the MAGA crowd want them to care!) - maybe wanting a public megaphone is more a thing that appeals to the other side.

Read the chess.com report. Hans admitted the 2020 cheating they detected.

Note that chess.com only accused (and got admission) of online cheating on their site. White noting irregularities of his recent win against Magnus, they are not accusing him of cheating, although they do note that in the post-game interview he was lying about his online cheating which prior to that that had been dealing with as a private matter.

I'd guess it means Musk had been told by his attorneys that he was almost bound to lose, so preferred to try to make it look like this was his choice rather than just being a loser coerced by the court.

It's interesting that he's not even attempting to offer any less than the original price he was on the hook for - maybe an indication of how strong Twitter's case was. The language is interesting though - "proposes to buy" vs "agreed to buy". Has anything changed in this new "proposal" ?

Got to wonder if the luke warm reception to TeslaBot, and/or Ukraine's "Fuck off" somehow have anything to do with it.

Sure, but 99.9% of medical students are hoping to become doctors, they are not going to end up in research careers designing new drugs. The few that do end up in research positions will surely be the most academically gifted who have no problem with organic chemistry or anything else.

That's all entirely reasonable, but I've got a slightly different take on it.

Families are paying astronomical amounts of money to send their kids to college in the US. An amount of money that can only be justified as the means to an end - getting that piece of paper that is (mostly needlessly) being asked for to get a decent job.

I've always thought that the *real* value of college, outside of a very small number of vocational degrees (medicine, law) is really as a halfway house between being looked after at home and being fully independent with a job, but that nice/useful phase of life experience is no longer justifiable as the main outcome given the cost. It's now all about the piece of paper (degree).

Of course you shouldn't be handing out passing grades entirely without justification, but I do think colleges have a responsibility to keep the material and grading at a level where the majority of students are able to graduate and therefore get a degree-based job. The role of universities in modern society isn't to weed out the smart kids from the less smart, but rather to provide further education and prepare kids for the workforce.

Of course there are kids who are more academically gifted than others, but we also supposedly have colleges along a spectrum of academic excellence who screen applicants based on GPA scores and academic capability (at least, this is what they should be doing). It seems quite reasonable to me that an MIT Computer Science graduate may be more capable (and more in demand from a jobs perspective) - part nature, part nurture - than one from, say, a poor quality community college, but OTOH I'd expect both to graduate a similar percentage of their students. It's not up to the colleges to set some absolute level of excellence necessary to pass courses - the level of material you can expect to teach will be a function of the quality of the institution and the quality of the students it is therefore able to attract. If you're a professor teaching at MIT then for sure set the bar high, knowing that the kids are up to it ... but you still need to grade on a curve to compensate for where you got it wrong as you're acting as gatekeeper rather than a functional part of society moving kids (of different abilities, from different colleges) from school into the workplace.