richardw

28 ms·

361.
by richardw 2y ago
I was repeatedly amazed at how smart my toddler was. You just feel the general intelligence. She's a bit older now (5) but e.g. a few days ago I was talking about cleaning the whole house. She said "you didn't clean the WHO
362.
by richardw 2y ago
I tried "are you sure", which often triggers some reasoning, and it was pretty confident. I'm trying not to give it the answer, but run it as if I didn't have any special knowledge. GPT + human > GPT. I mean, we'
363.
by richardw 2y ago
I know that's the messaging, but the real link to reality is very tenuous. This was a great example from the last couple days: https://twitter.com/colin_fraser/status/1785132544482226679 I just tried a simila
364.
by richardw 2y ago
Same thing with Facebook and identity. IIRC they leveraged Google’s address book to get traction, but will go after you if you try store FB social graph data long term for anything outside their garden. You try to block the tricks you used
365.
by richardw 2y ago
LLM’s are a compressed and lossy form of our combined writing output, which it turns out is similarly structured enough to make new combinations of text seem reasonable, even enough to display simple reasoning. I find it useful to think “wh
366.
by richardw 3y ago
It replaced “did you go to a top university” and “do you know our exact stack and tools” as the top hiring practice. Now anyone can compete for the same jobs from anywhere in the world. It’s a leveller, in my mind. I applied to Atlassian as
367.
by richardw 3y ago
OpenAI/MS are also building a DB of code recommendations and user fixes (GPT writes crap, you fix it), which is an excellent asset. I expect in 5 years it’ll be a couple of sentences to replicate most of whatever startups we create whi
368.
by richardw 3y ago
It diminishes the story that Databricks is the default route to privately trained models on your own data. Databricks jumped on the LLM bandwagon really quickly to good effect. Now every enterprise must at least consider Snowflake, and espe
369.
by richardw 3y ago
We had that same experience. My wife’s grandmother was 90-something when I got her an Android tablet thinking it would be easy to set up. She kept doing stuff that broke the experience for her. It would go home when she was trying to do som
370.
by richardw 3y ago
its partly in the article. Companies like Xbox and Lenovo want different experiences for their clients. This allows them to share the basic platform but specialise for their customer segments, eg professional or Xbox owners or whatever. Arc
371.
by richardw 3y ago
Recent immigrant to Australia here. Love the place, but the power of the gambling industry is alarming. Very interesting episode on the link with sports, and increasing earnings: https://www.theguardian.com/australia-news&#x
372.
by richardw 3y ago
The first job of an AI company is finding model/user fit.
373.
by richardw 3y ago
Not on premises who are regulated and might have enterprise requirements. If you need eg role based access control you can’t use on prem MLFlow etc. So you could niche down to customers who don’t know this yet. Smaller banks, insurers etc.
374.
by richardw 3y ago
There should be an open set of standards to allow users to select their search, travel, map, video, mail etc providers. You should get google mail results on bing search, Vimeo results on Google search, your travel bookings on any search en
375.
by richardw 3y ago
I was forced to orchestrate the replacement of Kafka with IBM MQ in a running financial institution because new execs understood it better from previous roles in other companies. Crazy newfangled micro services and event driven architecture
376.
by richardw 3y ago
Code. Every day. When you solve something tricky, you just basically released that as open source and trained ChatGPT 2030 how to do it without you.
377.
Generative AI to quantify uncertainty in weather forecasting(blog.research.google)
2 points by richardw 3y ago|0 comments
378.
by richardw 3y ago
It’s kinda simulating our brains but not really. When I attempted to dig more into how neurons work I realised that it’s a massive chasm of difference. Very much worth doing if you haven’t (you might know far better then me, this is for peo
379.
by richardw 3y ago
They do have a solid focus on doing so, it’s just not exclusive. https://www.databricks.com/product/machine-learning/large-la...
380.
by richardw 3y ago
I’ve wondered why email to print to post isn’t a bigger thing.
381.
by richardw 3y ago
I feel it’s a bit more, given we think about it after the sentence is complete. But it raises some interesting questions about what an agent would do if it had an instruction to keep trying until it got a reliable answer. Maybe an argument
382.
by richardw 3y ago
Most likely, unless you meditate a lot. Sometimes you'll take a bullet to save other people. Sometimes you'll drink yourself into a state that doesn't help you survive or reproduce. Or you'll write on a forum anonymously
383.
by richardw 3y ago
You haven't the vaguest fucking idea what I'm doing with the tools so put up a logical argument on the facts instead of just generating tokens in some way that attempts to play the person and not the ball. Here, argue with Yann, w
384.
by richardw 3y ago
I think so. You ask that question because you’re interrogating the position, not because 1000 humans have asked that question in similar situations. You and I know there’s a truth and we’d like to find it. The GPT is just happy (I.e. reward
385.
by richardw 3y ago
The fact that you ask that is in many ways the difference. You feel there’s a limitation in your knowledge of the term “understand” and its use in this context and would like clarification before you’re more certain, either way. At some poi
386.
by richardw 3y ago
You can't depend on it at all. I mean, you can use it for a tremendous amount of work, but until there is a way to constrain the bullshit LLM's can't be used for anything that requires a correct answer. The terms "depen
387.
by richardw 3y ago
But you’ve got 100 developers. That’s not “most web apps”, that’s firmly in the set of companies that need standardisation and potential scale, where the devops team makes the life of many devs far better. When it’s just a few devs and cash
388.
by richardw 3y ago
Very little to no evidence that continuously increasing intelligence will maintain the status quo indefinitely, on a scale of at least thousands of years. The odds of us not either populating the galaxy or utterly owning ourselves are vanis
389.
by richardw 3y ago
And AI is likely a binary outcome. It’s either very good for us or not, and it won’t take a million years to figure out which. We were in caves not long ago.
390.
by richardw 3y ago
If 10000 more kids shoot for the moon they’ll build a lot of skills that the equivalent lifers won’t. He’s saying things like: - Get good at tech - Build projects that you want - Get good grades and into the best university you can I reall

More ›