dougmwne

26 ms·

361.
by dougmwne 4y ago
Fair point. I didn't realize it was taxable as I am used to only profits being taxed and assumed Deepmind is run at a loss.
362.
by dougmwne 4y ago
People can be negative and critical for no reason. In this case, I think criticism is due because Google's failure to productize has lead them to a potential existential disaster. Most of their revenue depends on search and there being
363.
by dougmwne 4y ago
Github Copilot seems like a pretty clear example. They charge a subscription that's in excess of the marginal cost of inference.
364.
by dougmwne 4y ago
Piles of previous statements talking about safety and a general unwillingness to put any of its models in the hands of anyone not under NDA. Bard is the first counterexample I can think of and that was forced by OpenAI.
365.
by dougmwne 4y ago
Sounds like corporate funny money to me, not real revenue.
366.
by dougmwne 4y ago
Those are towering achievements that don't add to Google as a business. It was an important part of Google's reputation, but now that reputation is in the mud as Microsoft and OpenAI become king in the eyes of the public.
367.
by dougmwne 4y ago
That sounds like a fairly brilliant counter to "Open"AI. Something tells me that Google is still too scared of this tech in the hands of the public to go there though.
368.
by dougmwne 4y ago
I see your point, though I think it's ultimately going to be good for AI progress. So far the research has been mostly a vanity project for these companies. Who knew if there was really any gold at the end of those rainbows. Eventually
369.
by dougmwne 4y ago
Hah! Good catch. I doubt that was unintentional. Demis doesn't sound too pleased about having his merry band of misfits sucked into the mothership.
370.
by dougmwne 4y ago
Inventing transformers may prove to be Google's undoing. Being a research paper factory doesn't seem to be good for shareholders.
371.
by dougmwne 4y ago
PR blitz because it's going to look good to investors. Google is killing the vanity projects and moonshots and rolling those resources into teams that aim to launch products in the next 6-12 months.
372.
by dougmwne 4y ago
Yeah, absolutely lining him up to be the scapegoat. His chance for success seems severely compromised and his mission was always to invent AGI, not create some kind of lowly ad-search product. The party sounds over for him and I bet he wil
373.
by dougmwne 4y ago
While ultimately I think this is probably a very good organizational change, to have similar teams working on similar projects under the same leadership, it does seem to spell trouble in the short term. I can read between the lines that Goo
374.
by dougmwne 4y ago
It's a narrowing of the Overton window, not any one specific topic being accepted in public discourse.
375.
by dougmwne 4y ago
Anyone have a sense of just how much of a success or failure this was? THe way I understand it, getting the rocket off the launchpad was a major milestone. It seems they almost got to stage 2 separation.
376.
by dougmwne 4y ago
I wonder what happens if you just feel that dataset back into another LLM to re-write it and filter out the low quality items? IS there still any connection to the original copyright? How would that even be proven?
377.
by dougmwne 4y ago
The way the theory goes, people need to be encouraged to take risks. To gamble it all on a crazy new business idea that might not work out at all. Risk is the source of all new value. You can do this by making money very cheap and giving ve
378.
by dougmwne 4y ago
While I recognize the absolute importance of them, I'll set politics completely aside. Any idea how we could watch this with a dub? It sounds like a very interesting moment in film and I'd love to take a look.
379.
by dougmwne 4y ago
And it'll all be thanks to tightening monetary policy! Which is lighthearted, but not a joke. Economic theory says that the economic cycle helps to free up misallocated resources.
380.
by dougmwne 4y ago
The real thing we should be asking is if the 3B or 7B model is capable of correctly answering this question.
381.
by dougmwne 4y ago
It's not the JPG data you can copyright, but the recognisable image it produces. Every time you re-save or resize the image, the data changes, but the recognizable image remains. It's not clear how this process applies to model we
382.
by dougmwne 4y ago
https://github.com/tatsu-lab/stanford_alpaca
383.
by dougmwne 4y ago
The instruct tuning can be done with several open datasets at minimal cost. Should be easy for someone to create their own open model.
384.
by dougmwne 4y ago
I see this so often. People try to show what GPT can or can’t do without explicitly stating the version. Since not many people have access to GPT-4 it’s usually safe to assume they didn’t use it. GPT-4 is a massive improvement over 3.5 when
385.
by dougmwne 4y ago
This is absolutely huge. LLaMA weights leaking was a big deal, but you couldn’t actually use them without attracting Meta’s ire. Would love to see some benchmarking vs. LLaMA and GPT.
386.
by dougmwne 4y ago
While you are seeking out your expert friends and taking a long time, I’ll be on to the next 10 things I can learn from the AI.
387.
by dougmwne 4y ago
I had a similar experience recently having GPT-4 teach me double entry bookkeeping for a project I am ramping up on. I understood it in a vague sense from some Google, but wouldn’t be able to talk to a room full of accountants in an intelli
388.
by dougmwne 4y ago
Even when being queer is accepted, your life path is very different from your straight friends and family.
389.
by dougmwne 4y ago
I think the way you frame this is probably weird, but the actual behavior is completely normal. Most people would say that they care about their friends from previous phases of their life, but most social research says that friends rarely s
390.
by dougmwne 4y ago
This whole conversation about training set size is bizarre. No one ever asks what’s in the training set. Why would a trillion tokens of mundane gossip improve a LLMs ability to do anything valuable at all? If a scrape of the general interne

More ›