Something very interesting happened to me yesterday.
I'd been having conversations with ChatGPT about OpenClaw, nothing remarkable or extraordinary. Then I started a new conversation to talk about a different aspect, and GPT assumed I wanted to talk about some old PC game.
To disambiguate, I now had to refer to OpenClaw.ai. I asked it if it had some new system directive about this, and of course it denied it. Today we learn OpenAI has hired the OpenClaw developer, and he's "turning the project over to a foundation"
I believe that the rename to OpenClaw happened after the cutoff date for pretty much all model's training data, so unless you say something that causes them to look it up they'll get it wrong and assume it's about the old PC game. I was messing around trying to get different models to setup an OpenClaw NixOS VM for me and had to disambiguate for most of the models I tried.
Looks interesting, but the license is a show-stopper for me. I can't expose my employer to that kind of risk. I poured a lot of time and effort into learning Akka and has to throw it all away.
> I implemented a threaded interpreter [0] on it where the source was spread over 3 floppies. To assemble it I ran the 3 floppies through a small basic program that stripped all the comments out wrote it to a single floppy.
>
I had keyed the machine code to the CP/M version of figForth in on my Xitan Alpha. I sent it over a serial line to pip running on my Osborne after I bought one. Patched the console vectors and it was up and running.
> Back in the day I got to go to a demo and presentation about the machine.
>
> An interesting thing about the design was that it had a 5” screen, but it could only show about 50 columns at a time of the 80 column display. There was a knob you turned to pan the display.
On mine it's keyboard ctl-left/right arrow. Paint is worn off the metal surround as a result. I don't recall a knob.
The cost factor of international complications along with the impossibility of meeting every country's speech restrictions simultaneously may be why Elon is reducing Twitter's offshore presence dramatically. It could easily become a US-only platform. Which may be the only way for it to survive financially while meeting free speech goals.
Ya, it looked like a hard hit. But the announcer on the radio said "there go the retro thrust systems". So maybe they actually fired before impact and it just looked like an impact cloud?
Of course, she also called it an "off nominal situation" which sounds like a parody of corporate speak.
That cheary voice claiming the retro thrust success, turned me really off. There was no deceleration at all. All that dust was impact dust. It would have been much better to see that in silence, rather than hearing a forced attempt to feign success...
Frame by framing it it did look like it went off and sent out clouds before impact, but things were kind of behind a hill so it was a bit hard to judge (you can use ',' '.' to step forward and backward by frame on youtube desktop).
Rocketry traditionally contains a lot of euphemisms like "off nominal situation". E.g. the self destruct system on rockets is generally called a "range safety package".
I do feel like it's distinct from corporate speak though. Rocketry euphemisms are really just science-folks hedging their bets. Sure the rocket is doing something abnormal but the cause and result of it won't be properly clear until some research has been done so "off-nominal" seems like an apt description in the meantime.
I'd been having conversations with ChatGPT about OpenClaw, nothing remarkable or extraordinary. Then I started a new conversation to talk about a different aspect, and GPT assumed I wanted to talk about some old PC game.
To disambiguate, I now had to refer to OpenClaw.ai. I asked it if it had some new system directive about this, and of course it denied it. Today we learn OpenAI has hired the OpenClaw developer, and he's "turning the project over to a foundation"