Most of the conversations I'm talking about are not a quick whip out the phone and check...

Usually when I do this after the fact it comes back as "oh, that specific context wasn't given" or "oh, I misunderstood what you were asking", etc. Seems like the same...

Id say its accurate more than 90% of the time. I see erroneous facts added to the response that are wrong or not correctly represented the most.

Do you apply this same logic to the "experts" that you talk to? I talk with alot of SMEs from different domains. I could never fact check them all. Way too broad of subject matters. Similar to many of my indepth AI interactions

Id say its better than majority of my colleagues. Can you imagine if we fact check people in the same way?

Too some extent I agree with you. However, most of the time you really dont need to have "long" chats. In some ways I think some of the Flasehoods derive from the prompts themselves. Lots of bad/false/incorrect info is available on the web.

Can you imagine if we fact checked long conversations with people? How often do thing bad/false info would surface?

Im not an AI lover, but this critique that it is wrong 50% of the time is simply not true for newer models.

You've post 2 top level comments calling attention to the fact they are trans

not trying to be an asshole, but they're trans.

Naw. You are an asshole and trying to be.

I fully disagree that memory will be cheap in 2028. Please feel free to come and ping this thread in 2028 to tell me Im wrong.

My firm belief is that memory will continue to increase in price through the end of 2027 and likely level out in 2028. The price will not go 'down' in the next few years.

New technologies will be introduced, but all consumers and corporations want the latest and greatest. Demand will far outstrip supply for the foreseeable future.

Edit: just to add, if you keep up with the news and financial disclosurers, all the output from those new fabs has already been presold to the usual suspects.

Exactly. Expected to hit the market in 2027 and 2028. This investment in fabs happened years ago. For example, Samsung invested in the Taylor fab (which should start production, a little early, late this year) back in 2021. Samsung KRW6 was announced this year, but thats a new technology, not new capacity. You cant build and turn on a fab in the short term.

This is simply not true. All the major players are not investing in production. They are simply letting the demand stack and prices to rise.

China entering the market will help some, but expect these chips to be banned quite quickly in the western world for fear of communists. They are already blacklisted by many companies.

I pretty much only use prusament. Literally today I bought some overture for the first time. Read a bunch of posts and decided to give it a try.

I have to say this is really shitty reporting by Tom's and OPB. Sure its a big rate increase, but they left out the part where datacenters are still paying 30% to 50% less than residential customers per kWh.