Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
huijzer
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
59 ms
·
961.
▲
by
huijzer
4y ago
Basically every company is seeing a revenue decline in 2023 compared to last year. Even Apple reported a revenue decline of 5%. Meanwhile Microsoft is showing about 10% revenue growth in almost all their products, and the take-away is that
962.
▲
by
huijzer
4y ago
I always wonder how much I can trust settings like this, but this definitely looks like a step in a good direction!
963.
▲
by
huijzer
4y ago
There was a joke going around after the start of the war where someone, in response to a discussion about the effectiveness of face masks, responded with: "I‘m sorry, I‘m not an expert in virology anymore. Currently, I’m specialising i
964.
▲
by
huijzer
4y ago
I agree. I think the discussion has been in a better place. Many discussions are heavily polarized with no room in the middle. Maybe it’s the layoffs, maybe it’s the war, maybe it’s AI, maybe it’s the golem that Tim Urban talks about in his
965.
▲
by
huijzer
4y ago
> My main gripe with it though is that the actual interface is quite limited, and I don't like having to navigate to the site every time I want to use it. https://github.com/rikhuijzer/ata . Written in Rust. Doe
966.
▲
by
huijzer
4y ago
Oh. And also they are probably making ChatGPT Plugins ready for public release. Maybe the competition can catch up on the language model, but they will not likely catch up soon to the best language model with the most plugin integrations.
967.
▲
by
huijzer
4y ago
Although FOSS is great, extreme wealth inequality has to be fixed by the government and not by open source developers.
968.
▲
by
huijzer
4y ago
It could be that they are not training GPT-5 for a simple reason: Microsoft ran out of GPU compute [1] and they focus on meeting inference demand for now. Also, the GPT-4 message cap at chat.openai.com was shown as something along the lines
969.
▲
OpenAI has had to slow down development due to lack of GPUs
(theinformation.com)
2 points
by
huijzer
4y ago
|
2 comments
970.
▲
by
huijzer
4y ago
Summary: Trust us. We’ve been doing AI for years and are very good at it. We didn’t offer generative AI yet, so "today we are excited to announce Amazon Bedrock, a new service that makes FMs from AI21 Labs, Anthropic, Stability AI, and
971.
▲
by
huijzer
4y ago
Although a large total addressable market (TAM) is very alluring, know that most markets are dominated by a few players. For example, sugary beverages (Coca Cola), office software (Microsoft), or luxury sports cars (Ferrari). Exceptions are
972.
▲
by
huijzer
4y ago
Yes. No function overloading + no object oriented = verbose function names. What I find even more crazy is that some of the R functions take function names as strings as argument. For example, taking "sum" as an argument will then
973.
▲
by
huijzer
4y ago
This exact same difference exists between the North and the South of the Netherlands and also Germany. Even though it’s only a few hours away by car. In the US, the South is more explicit I guess? Here the North is more explicit.
974.
▲
by
huijzer
4y ago
Very interesting post. Sounds similar to the introduction of cars. Around the 1900s, there were hundreds of car manufacturers all jumping into this new market [1]. [1]: https://en.m.wikipedia.org/wiki/Timeline_of_motor_
975.
▲
by
huijzer
4y ago
In hindsight, the first quote already perfectly summarises what happens next: "Important public men like yourself are not 'common men.'" This binary thinking makes it a lot easier to throw people into the front line. Acc
976.
▲
by
huijzer
4y ago
Isn’t history full of examples of governments being slow and seemingly incompetent? Standard Oil was broken up many years after everyone knew that they were a ruthless monopoly which made too much profits. Note also that it didn’t take sena
977.
▲
by
huijzer
4y ago
For what it’s worth, I’ve been trying to verify the claim from Microsoft and OpenAI that we would get more software in the future when the price of software would decrease. In other words, they claim that there is a supply problem for softw
978.
▲
by
huijzer
4y ago
How about both? Both lax regulatory environment and expected AI productivity gains in the (near) future.
979.
▲
by
huijzer
4y ago
I couldn’t agree more. I listen to multiple audiobooks sort of intermittently plus multiple podcasts and it’s great to have such a mess. It’s about processing the information. What’s the point of reading a book if you don’t plan to remember
980.
▲
by
huijzer
4y ago
My favorite uses are: - Interactive debugging. Yesterday, for example, it helped me debug some tricky CSS issue where it gave hints and after 6 times back and forth, the solution came up. I had to explicitly set `-webkit-appearance: none` f
981.
▲
by
huijzer
4y ago
> It is the kind of innovation where anybody will catch up. Technically? Probably yes. But where is your startup going to find millions of users for extra training data?
982.
▲
by
huijzer
4y ago
> How much of Azure's staffing has been put on OpenAI projects? Great point!
983.
▲
by
huijzer
4y ago
Based on the speed at which OpenAI is shipping new products and assuming that they use their own technology, I'm starting to get more and more convinced that their technology is a superpower. Timeline of shipping by them (based on htt
984.
▲
by
huijzer
4y ago
> In fact it almost never had a chance, especially in the current environment where anticompetitive moves are no longer really pushed back on by the govt. They used to have a great market share of 100%, but then Microsoft bundled Interne
985.
▲
by
huijzer
4y ago
Even without Morgan Stanley claims, Greg Brockman and Satya have both said that GPT-3.5 was already internally used by them since the summer of last year and GPT-4 in December. Greg mentions it in https://youtu.be/YtJEfTTD_Y
986.
▲
by
huijzer
4y ago
> What does it mean to do it all as one step / how would I got about doing that with ChatGPT? Wait till they make the image input available via the API, I guess
987.
▲
by
huijzer
4y ago
Let me know if you got it working. I'm looking for such a thing too! Maybe stripping the styling and Javascript from webpages would work? Did you do the OCR as part of the complete model or did you make it a separate step? Machine lear
988.
▲
by
huijzer
4y ago
I think it can be okay in ecosystems where developers understand and respect semver
989.
▲
by
huijzer
4y ago
NuScale also still plans on a first reactor by 2027 even though are ahead in the small nuclear reactor space
990.
▲
Reboot for the AI Revolution by Harari (2017)
(nature.com)
1 points
by
huijzer
4y ago
|
0 comments
More ›