justanotheratom

you are assuming that all cancer research has been defunded altogether. And that if the government takes away some money, any ongoing research will come to a halt and we are doomed.

Also, maybe research dollars should emphasize prevention.

I don't enjoy the interface as such, rather the workflows that it enables.

I am sure this is worth every dime, but my workflow is so used to Cursor now (cursor rules, model choice, tab complete, to be specific), that I can't be bothered to try this out.

my question for anyone who knows:

Between SFT, DPO, and RFT, - when to use which? - can we mix and match? e.g, first SFT, then DPO.

honeycomb seems quite mature, thanks.

Great article and explanation.

On a practical level though, this would be the last thing I would use for log collection. I understand that when there is a spike, something has to be dropped. What should this something be?

I don't see the point of being "fair" about what is dropped.

I would use fairness as a last resort, after trying other things:

Drop lower priority logs: If your log messages have levels (debug, info, warning, error), prioritize higher-severity events, discarding the verbose/debug ones first.

Contextual grouping: Treat a sequence of logs as parts of an activity. For a successful activity, maybe record only the start and end events (or key state changes) and leave out repetitive in-between logs.

Aggregation and summarization: Instead of storing every log line during a spike, aggregate similar or redundant messages into a summarized entry. This not only reduces volume but also highlights trends.

well, you accuse me of lying, while you spew out nefarious theories.

I simply stated what the NIH director has publicly said. (https://www.nih.gov/news-events/news-releases/jay-bhattachar...)

RE: they'are increasing spending. - can you provide some evidence?

RE: their goal is to destroy things that serve the public... - you must be a genius to see through their goals. Has it occured to you that maybe they simply reflect that desire of the people you disagree with.

provide more facts and details.

Use to admire Gates, and good on him for doing what he is doing.

For me now, this statement by Larry Page resonates better:

“You know, if I were to get hit by a bus today, I should leave all of it to Elon Musk.”

This is a little too much fear mongering:

"excellent chance that you or a loved one will die from a potentially curable disease (cancer, dementia) because the research and clinical trials that should be bringing us those drugs are all being murdered."

Why is this an excellent chance?

Well, a straight answer that you may not like is that voters want different priorities: - reduce spending. - emphasize key public health priorities, like the chronic disease epidemic. - de-emphasize studies on gender identity, vaccine hesitancy, diversity, equity, and inclusion, climate change projects.

Yay! do you use your Gemini in Gemini App or AI Studio or Vertex AI?

I didn't mean generate separate TLA programs. Rather, other LLMs review and comment on whether this TLA program satisfies the user's specification.

Apple is generally on the side of customers, but this is a clear example of how anti-customer-friendly their policy was. As a customer, I had to jump through hoops to buy a book on their premium platform.

maybe run it through few other LLMs depending on how much confidence you need - o3 pro, gemini 2.5 pro, claude 3.7, grok 3, etc..

can websets enrich a column with images?

CloudKit not cross platform

You would be correct to be skeptical when they introduced AI into conversations, which btw is opt-in.

triplit sounds awesome. Any precedence of someone using it from a native iOS app?

"motive is to extract value out of their users who don't pay for a service" that is called a business.

if you find something deceitful in the business practice, that should certainly be called out and even prosecuted. I don't see why an effort to improve privacy has to get a skeptical treatment, because big business bad bla bla

I don't understand the knee-jerk skepticism. This is something they are doing to gain trust and encourage users to use AI on WhatsApp.

WhatsApp did not used to be end-to-end encypted, then in 2021 it was - a step in the right direction. Similary, AI interaction in WhatsApp today is not private, which is something they are trying to improve with this effort - another step in the right direction.

Supabase's decision to take a dependency on Deno, IMO caused indirect pain to lot of devs. I have wasted quite a bit of time trying to find or load a package that I needed. And now, with Deno 2.0 apparently everything is node compatible... I don't know what was the whole point.

is there a well-established tool-chain for finetuning these models?

FineTuning Ops platform requirement - compare evals, latency, cost across models.

Is Best-of-N Sampling standard practice these days in Inference? Sounds expensive on the face of it. I am surprised because I thought the trend was towards cheaper inference.

energy is abundant.

I feel like these issues can be countered by a reasoning AI that runs locally and I can configure to operate in my best interest.

e.g, Filter out political posts on X. Fact check opinion videos on the fly.

I hope future computing devices will have neutal engine at the center, and CPU as secondary. And I should be able to teach it to take actions on my behalf.

Congratulations on the launch. Suggest adding the demo video right in the landing page. That is the first thing I would be looking for. IMO.

Thanks for writing this. Suggest elaborating on the reason for using C++ in a Swift App.

This is quite cool. Two questions:

- why do you need nextjs frontend for what looks like a headless use case? - how much would be the OpenAI bill if there is 15 minutes of usage per day?