It is logically a 1080x1080 rectangular display. Every pixel outside the circle is ignored.
> as they don't seem as popular as search engines. Anecdotal evidence that I observe is that people stopped using google and only ask LLMs nowadays for their everyday queries.
LLMs are great for product research, as in "I want to buy X, what brands are there and what are the pros/cons and what do users report". Lots of potential to game the results there. If that is enough to sustain the…
So how about also getting rid of all regulations for wheelchair friendly infrastructure while we are at it? Way too expensive and it is even less than 2% of the population that requires it.
There are plenty of roaming applications: camping, boating, flying and war zones, which have both private and enterprise customer possibilities. Then there are remote places that either have no other option or would…
I don't think it is either of those. He is clearly not dumb nor do I think he is inherently evil. The thing is just that he became a billionaire at 23 and did not have to experience any of the grounding and setbacks…
Their target market is completely different. Anthropic and OpenAI try to build general AI that wins on all the benchmarks by throwing ungodly amounts of money at it. Mistral focuses on long term b2b contracts and their…
It is also valuable for scientists as it is often a 'directors cut' version of the paper. Journal submissions are heavy edited and shortened to fit into the page limits.
Tapping something is discoverable and a basic human instinct. How the hell do you discover "swipe up from the bottom edge and then slightly right" to open the app switcher?
Github Copilot lets you use Anthropic models. This is just a way for enterprises to fold this into their existing Microsoft contract instead of negotiating with a new vendor.
> LLMs can be described as "hardcoded weights for generating the next token". All code is "hardcoded". I disagree. Hardcoded means you need to change and recompile the entire program. So neither an .env file nor model…
I agree. I can tune a guitar to within a half a semitone of accuracy without reference just from knowing how the open strings are supposed to sound after 20 years of playing.
AI usage is very spiky and good models require very expensive hardware. Running locally would just result in it sitting idle ~90% of the time. I think renting will always be cheaper, for comparable performance at least.
Prime factorisation is not proven to be NP-hard so the existence of a quantum polynomial algorithm (Shor) has no bearing on complexity classes other than showing that prime factorisation is in BQP (quantum polynomial).…
> the game used several distinct pseudorandom number generators, to prevent e.g. randomness within a combat from influencing future card rewards. Why is this important? Feels like fixing what seems to be a non-issue…
Just look at the disruption smartphones and social media have caused. They had a gigantic impact to society and one definitely has to 'live with' that.
Absolutely not defending BAM or the police behaviour, but something that bothers me in the whole ordeal is that there is almost never any fact checking towards the other side. 1. The $200k figure seems is not grounded…
I would say his overall negative outlook is a well needed counterbalance to the completely delusional hype one is exposed to on a daily basis. The truth will then probably land somewhere in the middle.
I think the solution is not a dumbphone but a full android e-ink phone. Horrendous for everything that involves video but runs every app and can be used as an ereader. Unfortunately I have to found one that speaks to…
Because they lobbied for a rule change to get fast tracked to an index forcing passive investors to buy it at IPO price instead of being included after market value corrections.
I find it hard to see how that would ever be economical. LLMs need very expensive power hungry chips and datacenters have - bulk discounts - cheaper electricity - high utilisation to spread the costs among many users I…
He touches on this on the website and I agree with his statement > I do feel a lot of nostalgia for the days of trying to pack four people, four computers, and four monitors into one car on the way to a friends' LAN…
the era of subsidised ai is ending
One of Concordes problems was also that it was not that much faster for how uncomfortable it was. For London -> NY, You were looking at 7 hours in a luxurious business class vs 3.5 hours in a crammed noisy shaking metal…
Every company realising it is more profitable to milk their customers every day instead of just once.
It is logically a 1080x1080 rectangular display. Every pixel outside the circle is ignored.
> as they don't seem as popular as search engines. Anecdotal evidence that I observe is that people stopped using google and only ask LLMs nowadays for their everyday queries.
LLMs are great for product research, as in "I want to buy X, what brands are there and what are the pros/cons and what do users report". Lots of potential to game the results there. If that is enough to sustain the…
So how about also getting rid of all regulations for wheelchair friendly infrastructure while we are at it? Way too expensive and it is even less than 2% of the population that requires it.
There are plenty of roaming applications: camping, boating, flying and war zones, which have both private and enterprise customer possibilities. Then there are remote places that either have no other option or would…
I don't think it is either of those. He is clearly not dumb nor do I think he is inherently evil. The thing is just that he became a billionaire at 23 and did not have to experience any of the grounding and setbacks…
Their target market is completely different. Anthropic and OpenAI try to build general AI that wins on all the benchmarks by throwing ungodly amounts of money at it. Mistral focuses on long term b2b contracts and their…
It is also valuable for scientists as it is often a 'directors cut' version of the paper. Journal submissions are heavy edited and shortened to fit into the page limits.
Tapping something is discoverable and a basic human instinct. How the hell do you discover "swipe up from the bottom edge and then slightly right" to open the app switcher?
Github Copilot lets you use Anthropic models. This is just a way for enterprises to fold this into their existing Microsoft contract instead of negotiating with a new vendor.
> LLMs can be described as "hardcoded weights for generating the next token". All code is "hardcoded". I disagree. Hardcoded means you need to change and recompile the entire program. So neither an .env file nor model…
I agree. I can tune a guitar to within a half a semitone of accuracy without reference just from knowing how the open strings are supposed to sound after 20 years of playing.
AI usage is very spiky and good models require very expensive hardware. Running locally would just result in it sitting idle ~90% of the time. I think renting will always be cheaper, for comparable performance at least.
Prime factorisation is not proven to be NP-hard so the existence of a quantum polynomial algorithm (Shor) has no bearing on complexity classes other than showing that prime factorisation is in BQP (quantum polynomial).…
> the game used several distinct pseudorandom number generators, to prevent e.g. randomness within a combat from influencing future card rewards. Why is this important? Feels like fixing what seems to be a non-issue…
Just look at the disruption smartphones and social media have caused. They had a gigantic impact to society and one definitely has to 'live with' that.
Absolutely not defending BAM or the police behaviour, but something that bothers me in the whole ordeal is that there is almost never any fact checking towards the other side. 1. The $200k figure seems is not grounded…
I would say his overall negative outlook is a well needed counterbalance to the completely delusional hype one is exposed to on a daily basis. The truth will then probably land somewhere in the middle.
I think the solution is not a dumbphone but a full android e-ink phone. Horrendous for everything that involves video but runs every app and can be used as an ereader. Unfortunately I have to found one that speaks to…
Because they lobbied for a rule change to get fast tracked to an index forcing passive investors to buy it at IPO price instead of being included after market value corrections.
I find it hard to see how that would ever be economical. LLMs need very expensive power hungry chips and datacenters have - bulk discounts - cheaper electricity - high utilisation to spread the costs among many users I…
He touches on this on the website and I agree with his statement > I do feel a lot of nostalgia for the days of trying to pack four people, four computers, and four monitors into one car on the way to a friends' LAN…
the era of subsidised ai is ending
One of Concordes problems was also that it was not that much faster for how uncomfortable it was. For London -> NY, You were looking at 7 hours in a luxurious business class vs 3.5 hours in a crammed noisy shaking metal…
Every company realising it is more profitable to milk their customers every day instead of just once.