Ratelman

What do you mean when you say sexual evolution?

"The “after” curve is higher than the “before” curve to the right of 140ms which means more requests are finishing slower than before."

Think they made a typo - "before" curve is higher than the "after" curve to the right of 140ms, which means more requests take longer than before to finish.

Or am I the one misunderstanding?

I actually like this idea - makes sense at face value - as long as they design the test in such a way that it aptly applies the knowledge instead of just learning for the sake of passing test like questions...

I get where you're coming from, was shocked when I left a relatively well organised corporate and did work at a relatively older company with a ton of legacy systems - when I asked what the strategy was they explained the structure to me - at the year end results they highlighted that they hit targets of cost cutting and saw this as an achievement, the whole narrative was around how its a tough economic environment (the presentation was literally all about things happening in the world - nothing about things they did/projects they delivered/value they added...) - they also had more project managers than engineers and wondered why projects kept missing deadlines- they hoped AI would solve their problems - but you can't get ROI in a space like that where your engineers are using AI to patch the ship to keep it afloat while the project managers think they're in an airplane and are trying to get it off the ground...

A financial services company in Africa

I've seen legitimately good outcomes with AI - a backlog has been cleared, features that were left on the cutting room floor have been pulled back in AND delivered all thanks to the use of AI coding tools. AI workflows have brought down processes from weeks of human processing to a couple of minutes with human oversight - and the revenue that it unlocks more than covers the AI bill. This is within a large corporate company - the "No such story exists for AI" feels overplayed. Sure, the wave of (quoting the article) "braindead executives, imbeciles and middle management hall monitors that don’t do any real work" might be bigger than with previous hype cycles because AI as a tool does enable pseudo-intellectualism, but the article overstates its case. I know, 1 counterpoint doesn't make a strong argument - but there's no reason the way we're applying this as a tool can't provide the same gains within other organisations - am I missing something/being delusional/huffing copium?

If you are eyeing the South African market - I can promise you granting credit here is waaaaayyy ahead of the US. There is a very solid credit bureau and a few of the banks are already on the "use AI to process docs" train. For rest of Africa - they're bigger on using cellphone data (see Optasia). If you want some insight into the market - happy to have a chat (email on profile)

Reflecting on my own experience - frequency of contact (if I see them once a year, can't really count them as close friends) How involved they are in my life - are they people I turn to when I'm facing a problem, do they turn to me when facing their own problems? Do we have frequent deep conversations - not just surface level discuss the weather, sports etc. but stuff that matter. Quantifying this - length of friendship (# of years), frequency of contact (annually, monthly, weekly etc.), level of trust (low, medium, high - can I trust my kids with them kind of trust), level of involvement (low, medium, high - what things do I feel comfortable sharing with them - suppose this is also level of trust?)

What makes you consider them close (aside from length of friendship)?

Boils down to the basics of proper science - how does one measure/quantify close friends?

Yeah, he was quite vocal in his opinion that they would plateau earlier than they did and that little value would be derived from them because they're just stochastic parrots. Agree with him that they're probably not sufficient for AGI, but, at least in my experience, they're adding a lot of value and they're continuously performing better in a range of tasks that he wasn't expecting them to.

Was my thinking exactly - but also semantically equivalent is also only relevant when it needs to be factual, not necessarily for ALL outputs (if we're aiming for LLM's to present as "human" - or for interactions with LLMs to be natural conversational...). This excludes the world where LLMs act as agents - where you would of course always like the LLM to be factual and thus deterministic.

In a few years we've gone from gibberish (less poetic maybe, less polished and surprising, but none the less gibberish) - to legit conversational, and in my own opinion, well rounded answers. This is a great example of hard-core engineering - no matter what your opinion of the organisation and saltman is, they have built something amazing. I do hope they continue with their improvements, it's honestly the most useful tool in my arsenal since stackoverflow.