Its a great shame the imbalance. Ted from Scrubs never got near the same adulation that Chandler from friends got when passing. Ted was up there with our own Ted as sitcom greats . I only mention this as I saw a lookalike today of our balded hero and took me right back.
I thought I might start a history substack, writing short pieces on interesting events that are relatively unknown. The first piece is on the Irish & German mercenary revolt in Rio, does anyone fancy giving it a quick read and giving me some feedback?
Originally posted by Hitchhiker's Guide To...View Post
Interest rates now expected to start falling in April. An element of relief for those of us coming off fixed rates next year and who don't want to go through the hassle of switching mortgage providers.
Read similar earlier, I wonder if we’ll see much a divergence between EU & US over the next couple of years.
On RD3 getting his Atlantic sub, that link is a great value sub at 1e a month.
I thought I might start a history substack, writing short pieces on interesting events that are relatively unknown. The first piece is on the Irish & German mercenary revolt in Rio, does anyone fancy giving it a quick read and giving me some feedback?
shrapnel any food tips for Lyon? Would assume eating something authentic like a pigs head is de rigeur?
Sadly not. Never spent any time there, but definitely has the reputation as one of France's top gourmet destinations, so I'm sure there'll be loads of phenomenal options
Had forgotten about that one - the face on him as he starts walking off and realises that actions have consequences...
One of the best meals I ever had was in a tiny chef patron place somewhere up near Croix-Rousse that I stumbled into by chance... I can't locate it on google maps, I guess it was over 20 years ago so it's probably long gone.
City is teeming with great food, will be hard to go wrong.
Stick away from quinelles. They are brutal. You'll be tempted, because they are advertised absolutely everywhere as a classic local food, but then you'll be disappointed, and then you'll think of me saying how disappointed you would be, and then you'll think - maybe hitch is right on everything else, and thats not a good place to be at in life.
I was using the advanced data analytics plugin a few days back to help draw up something on which I have very little knowledge. I had it cross referencing its own knowledge on the subject along with two pdfs I fed it with extra information and another pdf with guidelines and constraints. Its quite clunky and sometimes the ways in which it extracts information from the documents by doing some kind of brute force searching seems off as its not using the information in the context of knowledge but more a text it can look up which isn't ideal but I got some amazing results when it worked where it was cross referencing all the info and coming up with some great output. It is such a crazy way of doing something where you feel almost like you are directing this tool to create something and you end up in this kind of collaborative effort. If I could directly train it on these added pieces of data with enough weight that it holds them in some kind of higher context for the conversation but it is part of its training data rather than whatever way the plugin is handling it then that would be incredible.
They did a Keynote speech today with major updates and it looks like this is one of the major features they will be rolling out over the next few weeks. You can basically create a Custom GPT where you upload additional data that you would like the modal trained on.
OpenAI developers conference today is all manner of ridic progress. You can, from next month, build your own personal model, or loads of different personal models, depending on what you want to be advised on. So you'll build up these private models and have them as these perfectly trained assistants for anything you want but only you will have them. We're going well past the looking glass here.
There's a bit of an intelligence stretch needed in GPT4 to get it quite there, but its already astonishingly close to something that resembles supreme intelligence. Like, GPT5 if its maybe even just 50% more intelligent - assuming thats possible - will be, by a long shot the most intelligent person alive at many many many disciplines.
And it can be 200 million workers just as easily as it can be 1 worker, for a fraction of the cost.
The Chinese seem to have caught up also with Ernie 4.0, which appears to be capable of generating fully fledged videos from prompts. Although am not sure if anyone has actually used it hands-on yet. Then Google Gemini which appears like it will be GPT 4.25 at least and is soon to be released.
I think the extent to which we haven't seen anything like this before can't be understated. Like, our entire concept of white collar work is going to completely collapse within five years. They'll be keeping us on as charity cases while the algorithms do everything. Or maybe we just won't have white collar jobs.
Originally posted by Hitchhiker's Guide To...View Post
ha. we have very similar interests OPR!
I remember right around when GPT first really entered the public consciousness and my own that after playing with it for a while, my mind jumped right to the end game of thinking, what is the point? The evolution of this to where it ends up completely changes everything. I see all the posts on reddit complaining about how this particular update destroys all their hard work and effort. We are in this weird stage where the technology allows people to build and create awesome things they never could before but the landscape doesn't stay static for long enough to be useful for them and instead it is advancing at such a rate that eventually the tech eats any innovation that others have made outside it.
I keep hearing people talking about how they have this window of time until more people start using AI and they need to use this competitive advantage for as long as they have it to get ahead. The thing is if AI advances as we think it will in the coming years it makes almost everything obsolete. At some stage the systems will be so much better that it will create newer and better versions of anything humans can currently do in conjunction with it. Sam even ended the keynote by saying these updates are nothing compared to what they are currently work on.
This is so different from anything that has come before that people don't seem to be able to conceptualize and internalize that fact. It is also quite nihilistic in nature because you are essentially saying all effort is futile.
I remember right around when GPT first really entered the public consciousness and my own that after playing with it for a while, my mind jumped right to the end game of thinking, what is the point? The evolution of this to where it ends up completely changes everything. I see all the posts on reddit complaining about how this particular update destroys all their hard work and effort. We are in this weird stage where the technology allows people to build and create awesome things they never could before but the landscape doesn't stay static for long enough to be useful for them and instead it is advancing at such a rate that eventually the tech eats any innovation that others have made outside it.
I keep hearing people talking about how they have this window of time until more people start using AI and they need to use this competitive advantage for as long as they have it to get ahead. The thing is if AI advances as we think it will in the coming years it makes almost everything obsolete. At some stage the systems will be so much better that it will create newer and better versions of anything humans can currently do in conjunction with it. Sam even ended the keynote by saying these updates are nothing compared to what they are currently work on.
This is so different from anything that has come before that people don't seem to be able to conceptualize and internalize that fact. It is also quite nihilistic in nature because you are essentially saying all effort is futile.
Yeah, its genuinely mad what is happening. I do think though theres a bit of time before it quite seeps through, and that bit of time can be valuable if used right. Its not much time though. But maybe doing something now unlocks the ability to do something even bigger in the future.
Originally posted by Hitchhiker's Guide To...View Post
Yeah, its genuinely mad what is happening. I do think though theres a bit of time before it quite seeps through, and that bit of time can be valuable if used right. Its not much time though. But maybe doing something now unlocks the ability to do something even bigger in the future.
The time horizon could still be decades. I'd probably bet on timeframes closer to 5 years though given the people closest to the tech seem to think in these terms. You don't even need AGI but just an advanced system that can act almost indistinguishable from it by solving practical problems like working memory, context, etc. Even at that stage if we don't term it to be alive you get into all sorts of mad questions around what is consciousness. If you listen to enough Sam Harris around where thoughts come from and what is consciousness you begin to wonder if we are some form of large language model
At some point though your knowledge and ability gets superseded by the machine which is where it all breaks. Your ability to do something bigger in the future doesn't apply because you have something more intelligent and better that you now wondering what it can do with the same information.
One of the best meals I ever had was in a tiny chef patron place somewhere up near Croix-Rousse that I stumbled into by chance.
Thinking back to that meal now, I was wandering around Croix Rousse on my own on a Sunday night and liked the look of this place , very small very traditional, I got the only free table in the place. I was given a menu that I couldn't even begin to translate so looking at the table beside me where two people were in the middle of their main, one was a delicious looking fish dish and the other was a revolting looking home made sort of sausage - lumps of all sorts of innards roughly chopped and stuffed inside an undisguised bit of intestine that looked suspiciously close to the rectum. Obv I pointed at the fish dish but after a fantastic starter of gazpacho I was presented with the vile looking sausage, I tried not to look at it and poked around a bit so as to have at least made some impression of having tried it. It was incredible! how could something so repulsive looking taste so good. A simple tarte tatin for desert has never since been matched.
Originally posted by Hitchhiker's Guide To...View Post
I think the extent to which we haven't seen anything like this before can't be understated. Like, our entire concept of white collar work is going to completely collapse within five years. They'll be keeping us on as charity cases while the algorithms do everything. Or maybe we just won't have white collar jobs.
I wonder what this means for future business. Might the advice of investing in index funds be a thing of the past?
I've seen a bunch of indie developers make millions from wrapper apps. But it seems a futile game because everything moves so fast, and what about the rest of day to day business. IP and tech might mean very little in a time when anyone could conjure up any software they want.
Big tech won't be gate keepers any more. The value in markets may swing back towards business that do tangible things; while tech goes to nothing. Amazon, Microsoft, Google may have to settle for being the IBM of old and making their money from hardware via renting cloud servers.
I barely use Google any more. Much better and easier to use chatgpt. I've made a few personal apps using chatgpt4. As that gets even better, why would I ever need to buy off the shelf apps or software ever again?
Maybe the future is investing in companies with moats in highly regulated industries or capital intensive areas. FANG will be replaced with anyone that has access to chatgpt.
The end or the begining is nigh. We are all chatgpt'd up. The missus conjured up a report using chatgpt and informed the work audience of her source . All were suitably impressed and wary .
Come gather round people wherever you roam and admit that the waters around you have grown.
There's always millennials complaining that they followed the advice, got good results, went to college, worked hard, but they can't get any good jobs. Maybe the next version is: I invested for 30 years and now the NASDAQ is worth zero because anyone with a phone can be FANG and I can't retire
I thought I might start a history substack, writing short pieces on interesting events that are relatively unknown. The first piece is on the Irish & German mercenary revolt in Rio, does anyone fancy giving it a quick read and giving me some feedback?
I enjoyed that. Short and to the point while telling the story well. Made better by the fact I'd never heard of the story. 5 wombats.
It can't even get Ohm's Law right in the first place, so I haven't gone there!
Is this wrong?
This was after giving it suitable context to answer the question, which tends to have a big impact on quality of output. The context was:
You are a world-renowned Physics Professor who is giving a guest lecture to undergraduate physics students. The topic is Ohm's Law. Please provide a summary of the law as you would explain to these students. Think about this step-by-step: are you explaining all necessary information, and at the right level for physics undergraduate students.
Response:
Ohm's Law is a fundamental principle in the field of electronics and electrical engineering, describing the relationship between voltage, current, and resistance in an electrical circuit. It’s named after the physicist Georg Simon Ohm.
The law states that the current flowing through a conductor between two points is directly proportional to the voltage across the two points, provided the temperature remains constant. This can be mathematically expressed by the equation:
\[ I = \frac{V}{R} \]
Where:
- \( I \) is the current in amperes (A)
- \( V \) is the voltage in volts (V)
- \( R \) is the resistance in ohms (Ω)
To understand Ohm's Law, let's break down each component of the equation:
1. **Current (I):** This is the flow of electric charge in a circuit, measured in amperes. It can be thought of as the flow rate of the charge. In more physical terms, it's the number of electrons passing through a point in the circuit per second.
2. **Voltage (V):** Voltage, also known as electric potential difference, is the force that pushes the current through the circuit. It's the energy per unit charge exerted by an electric field on a charged particle.
3. **Resistance (R):** Resistance is the opposition that a substance offers to the flow of electric current. It’s like the friction that opposes the flow of water through a pipe. The higher the resistance, the lower the current for a given voltage.
Ohm's Law also allows us to understand how the components in a circuit behave. For example, if we increase the voltage, the current will increase, provided the resistance stays the same. Conversely, if we increase the resistance, the current will decrease if the voltage is kept constant.
In practical terms, when you're designing a circuit, you might use Ohm's Law to determine what kind of resistor you need to get the desired current flow with a given voltage supply. It's also used to calculate power dissipation with the formula \( P = VI \), which is derived from Ohm's Law.
Understanding Ohm's Law is critical for anyone working with electrical circuits, as it applies to direct current (DC) and, under certain conditions, to alternating current (AC) as well. It's one of the first and most important laws you'll learn in physics, especially as you delve into electronics.
I think its safe to say is that the reason why Zuut hasn't responded is because he is checking his contract to see if he can legally be replaced by ChatGPT.
Originally posted by Hitchhiker's Guide To...View Post
I think its safe to say is that the reason why Zuut hasn't responded is because he is checking his contract to see if he can legally be replaced by ChatGPT.
You haven't posted any nonsense today. HitchGPT assumed full control?
"We are not Europeans. Those people on the continent are freaks."
Thinking back to that meal now, I was wandering around Croix Rousse on my own on a Sunday night and liked the look of this place , very small very traditional, I got the only free table in the place. I was given a menu that I couldn't even begin to translate so looking at the table beside me where two people were in the middle of their main, one was a delicious looking fish dish and the other was a revolting looking home made sort of sausage - lumps of all sorts of innards roughly chopped and stuffed inside an undisguised bit of intestine that looked suspiciously close to the rectum. Obv I pointed at the fish dish but after a fantastic starter of gazpacho I was presented with the vile looking sausage, I tried not to look at it and poked around a bit so as to have at least made some impression of having tried it. It was incredible! how could something so repulsive looking taste so good. A simple tarte tatin for desert has never since been matched.
Sounds like andouillette, which I also once mistakenly ordered. A fine tang of what can only be described as piss.
"We are not Europeans. Those people on the continent are freaks."
Originally posted by Hitchhiker's Guide To...View Post
I think its safe to say is that the reason why Zuut hasn't responded is because he is checking his contract to see if he can legally be replaced by ChatGPT.
I was following up a 2-week, €10k repair to a pump that took 10 weeks and it still doesn't work. When I queried this point, I was told 'it was fine when it left the factory'
What you need for sim card type sniffing detection is a troll. I don't know if any of you saw the Swedish film Border from 5 years ago about a couple of trolls.
Haven’t seen it. But it reminded me of the Norwegian film on Netflix. Troll
Originally posted by Hitchhiker's Guide To...View Post
Lets not forget the capacity of Scooby Do for sniffing out arch-villains.
I was reading something about a recent case in New Zealand, but can't remember the details. Someone who was imprisoned for decades and ended up innocent. The point of the thing I was reading was that there was someone else in a New Zealand prison also for murder who had an even weaker case against them, but their problem was that they 'looked' like the type of person who would carry out the particular murder they were accused of.
I wonder with this Ozzie lady does she face a problem that all of our first thoughts on seeing her were: yep, you look exactly like thats what you would do - frumpy, ditched, presumed very very bitter.
Like, just to break it down on the surface: she apparently had this plan to poison family members by finding deadly mushrooms but only thought to get rid of the food dehydrator after the fact? It seems a lot more like the type of thing you would do if felt you were being suspected, rather than if you were actually intending to commit murder.
The area has deadly mushrooms - I'd assume every area has deadly mushrooms. I went through a phase of mushroom foraging and had to give it up as you learn fairly quickly that the default assumption is that every mushroom is deadly.
She claimed to keep the mushrooms in her cupboard but couldn't remember where they came from - we have dried mushrooms in our cupboard and can't remember where they came from.
That’s a nice version of events. But there there are a few holes in it.
I heard the details of the case before I seen a picture of her. And when I did, the fact she was “frumpy” didn’t influence how I view the case. A bit disappointing that you’d think, and think everyone else agrees.
There’s no evidence she got rid of the dehydrator after he fact. Could have been before she cooked meal. And only invent the timeline after her ex-husband brought up the dehydrator he had previously.
Deadly mushrooms are very common. But it was specifically the type that they were poisoned with that grew in the area. And even if death caps are ubiquitous, that doesn’t explain in the slightest how they got into the Wellington.
She didn’t claim she got them from the cupboard. She claimed she got them from an Asian shop, but couldn’t remember where. It’s a rural town of 5000 people. I highly doubt there are a multitude of ethnic shops nearby.
And sure, you may not remember exactly where you got your porcinis at the back of the cupboard. That’s not strange. But it’s pretty important to point out that they are probably not dried death caps.
It may have been an accident, but I have actually heard of any plausible explanation for how death caps got into the beef Wellington. She hardly forgot to mention that she recently decided to give foraging a go
I’ve said I wouldn’t make my mind up, until all the details come out - including the evidence that led to multiple charges.
But for the sake of balance, let’s not forget the case of Lindy Chamberlain, aka “a dingo ate my baby” lady falsely accused and convicted of murdering her baby.
I thought I might start a history substack, writing short pieces on interesting events that are relatively unknown. The first piece is on the Irish & German mercenary revolt in Rio, does anyone fancy giving it a quick read and giving me some feedback?
ChatGPT Question: currently interviewing for a position in our firm, sent all decent applicants a task to write two separate one page summaries on specific things. One needs up to date information and one is more generic, would you allow an obvious ChatGPT response to go through?
ChatGPT Question: currently interviewing for a position in our firm, sent all decent applicants a task to write two separate one page summaries on specific things. One needs up to date information and one is more generic, would you allow an obvious ChatGPT response to go through?
ChatGPT default answers tend to be excessively wordy and generic, so if someone just runs it through ChatGPT and gets that and then sends it to you I'd simply judge them on not submitting a great answer to the request that you set.
I heard the details of the case before I seen a picture of her. And when I did, the fact she was “frumpy” didn’t influence how I view the case. A bit disappointing that you’d think, and think everyone else agrees.
I find this very unlikely. How could you possible have heard the details within seeing a picture of her - there would be internet story, newspaper story, tv story that wouldn't include her picture. If you heard it secondhand then you heard it from someone who was possibly judging her from her appearance.
But for the sake of balance, let’s not forget the case of Lindy Chamberlain, aka “a dingo ate my baby” lady falsely accused and convicted of murdering her baby.
Yeah I might be giving quite a bit of weight to this - of assuming that generally Australians jump to conclusions in legal matters, and then follow through on them out of pure pig stubbornness and ignorance.
Now obv I haven't a clue what the truth in this particular circumstance is, I was really just remembering that New Zealand story I read which was making the point about someone being judging as looking like someone who would carry out a particular crime and as a result being judged more quickly than someone else who didn't look the part - even though the evidence was weaker in the case of the person who looked the part. I'm fairly certain it was the Mark Lundy case: https://en.wikipedia.org/wiki/Lundy_murders
EDIT: just reading the Lundy case in that wiki link above. wowsers its more brutal than I remember.
Think it was a video I first saw the story in, about his first appearance in public where everyone judged that he appeared to be over-acting, which was then argued to have followed him through the criminal system with a presumption of guilt based on first impressions.
In terms of reading about GPT in a practical what does it mean sense, I'd recommend Ethan Mollick over anyone. He's supremely gifted in working out what GenAI means in practice: https://www.oneusefulthing.org/
Back in surgery tomorrow morning for the big one. If all goes well, this is the start of the growing/healing process. It's only taken 1.5 years to get here no thanks to the NHS (only 3 months of that in the Swiss system), who funnily enough called me this afternoon for surgery pre assessment, yet when asked, didn't have a surgery date on file. i've had 3 or 4 pre assessments with them, it's their way of stringing you along, giving you hope.
Originally posted by Hitchhiker's Guide To...View Post
ChatGPT default answers tend to be excessively wordy and generic, so if someone just runs it through ChatGPT and gets that and then sends it to you I'd simply judge them on not submitting a great answer to the request that you set.
To me using it for research is fine, but copying and pasting the answer is ridiculous. It’s very obvious and I spotted it after the first one, there were about 30% responses from it which I was disappointed by,
Got my first ever speeding fixed penalty notice today!
30 years driving and they finally caught me
59KPH in a 50
Its a 200 metre stretch on the Firhouse Road that I was on
Usually drive at prevailing speed which is usually slow (but not slow enough apparently) as it's a small stretch before you're turning right. I actually feel like I'm going too slow in this stretch most of the time.
Huge wide road too
Fair enough but it's going to be murder keeping to even lower limits with everyone driving up your hole
Im mostly a slow enough driver no interest in the overtake lane etc. Now I'm going to have stress out in all these zones as I'm going to be the slowest fucker there
To me using it for research is fine, but copying and pasting the answer is ridiculous. It’s very obvious and I spotted it after the first one, there were about 30% responses from it which I was disappointed by,
I wonder if you’ll see fewer companies send out case studies and coding exercises given it’s prevalence.
We don’t do any take-home anymore, I believe we did in the past but it’s all live SQL/R/Python & case studies.
Got my first ever speeding fixed penalty notice today!
30 years driving and they finally caught me
59KPH in a 50
Its a 200 metre stretch on the Firhouse Road that I was on
Usually drive at prevailing speed which is usually slow (but not slow enough apparently) as it's a small stretch before you're turning right. I actually feel like I'm going too slow in this stretch most of the time.
Huge wide road too
Fair enough but it's going to be murder keeping to even lower limits with everyone driving up your hole
Im mostly a slow enough driver no interest in the overtake lane etc. Now I'm going to have stress out in all these zones as I'm going to be the slowest fucker there
Hope the insurance doesn't go up next year
€160 fine ..balls to it
Got our first speeding ticket here as well. 46 (readjusted to 41 as per margin of error rules) in a 40
Apart from the shame you've brought on your family, I think you might be grand on insurance. Heard its only when theres a second set of points that the insurance companies pay any real notice to it in terms of bumping up rates.
Greatest ever innings by maxwell in that game. Hard to believe. Aus 91 for 7 chasing 290 or so. He then goes on to hit 200 whilst not being able to move due to cramp (barely any runs other than 4s or 6s).phenomenal
To me using it for research is fine, but copying and pasting the answer is ridiculous. It’s very obvious and I spotted it after the first one, there were about 30% responses from it which I was disappointed by,
Are you using the paid version or the free version? The free ones hallucinates far more than the paid one. I know for coding the difference between the two is night and day.
There might be a plugin which makes it better for your specific questions. They seem to be accessible within the store already. They have basically nuked any of the decent wrappers with in house versions.
Like for instance I see someone on reddit that gave the data analyzer this engineering drawing and from the picture it is able to produce the below output. I think they wanted to go from drawing to list of components.
ChatGPT Question: currently interviewing for a position in our firm, sent all decent applicants a task to write two separate one page summaries on specific things. One needs up to date information and one is more generic, would you allow an obvious ChatGPT response to go through?
Working on the presumption that the task is vaguely related to the day to day. Big stretch based on some companies interview processes.
Would an obvious chat gpt response be fine in your day to day or would that be a problem?
It’s a pretty big shortcut to take so do you want someone who will look for shortcuts to speed delivery or do you need someone who will sweat the details?
Subjective based on what you need to fill would be my opinion.
I just use the free version. I have no interest in AI really. Should be able to hang on for 15 years in dinosaur mode given how Universities work, then off into the sunset!
We process personal data about users of our site, through the use of cookies and other technologies, to deliver our services, personalize advertising, and to analyze site activity. We may share certain information about our users with our advertising and analytics partners. For additional details, refer to our Privacy Policy.
By clicking "I AGREE" below, you agree to our Privacy Policy and our personal data processing and cookie practices as described therein. You also acknowledge that this forum may be hosted outside your country and you consent to the collection, storage, and processing of your data in the country where this forum is hosted.
Comment