Amazing that it’s impossible to tell whether you are a real crypto booster or joking at their expense.

> "[...] Thanks for the great ride!" Vogt wrote.

Not the best phrase to use in response to this scenario...

It’s possible to prove.

Use an LLM to do a real world task that you should be able to achieve by reasoning.

With only the information we had in 2020, the two theories “language models don’t reason, they regurgitate” and “as language models scale, they begin to think and reason” made predictions, and the people who invested time and money based on the predictions of the latter theory have done well for themselves.

Where is the state of the art at with self-driving two wheeled vehicles?

All of you criticizing the mismatch of power between a Vespa and a tank simply do not understand the desperation felt by those living in countries occupied by the Germans during the war. In the 1950s Europe was poor, extremely traumatized, and people lived in fear of another war. This is a vehicle of desperation.

I really really want this to work.

However at this point - benchmark success is about as effective as results from someone who has been “taught the test”

If say… Merck wanted to use this same model to reason out a logistics issue, or apply it to some business problem at scale - you’d have to deal with hallucinations all over the place.

The best analogy I have right now is that improved results on benchmarks are like better acting from Hugh Laurie as House.

If you want to watch a show - great (generative work)

If you want to get a prescription - then not so much.

Swimsuit or not, people can behave inappropriately. Let me put it this way, the couple were having some form of sex and thought they were being discreet. The acoustics betrayed them.