TheFullStackSnack

886 posts

TheFullStackSnack banner
TheFullStackSnack

TheFullStackSnack

@f_stack_snack

20+ years in programming for fun and profit | Improving tech one snack at a time 🍪💻 | working on local-first log querying tool 📊

Katılım Ekim 2024
66 Takip Edilen30 Takipçiler
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
Using chatgpt+aistudio as a daily driver. Decided to give @claudeai a try with $20 plan an opus 4.7. Here are quotes from a single conversation. How do people find this usable?
TheFullStackSnack tweet media
English
0
0
0
30
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
I feel they need checkboxes there for more complete feedback
TheFullStackSnack tweet media
English
0
0
0
3
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
@claudeai is there a way to have yolo mode for dispatch? I have it running on a dedicated VM, that can be recreated in minutes with its own accounts. I want it to be helpful and autonomous when I send tasks from phone, but I have to stare at it and click allow every 15 seconds.
English
0
0
0
7
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
It only took 9 years for JetBrains to get to my issue… by a bot. What a time to be alive =)
TheFullStackSnack tweet media
English
0
0
0
57
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
I guess @sama decided that having Anthropic at #1 on the App Store wasn't enough, so he decided to alienate developers too. He’s made Codex so dumb today that I’ve triple-checked all my settings, and now I’m back to typing code manually since it’s hopeless.
English
0
0
1
20
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
@OpenAI Can we please don't degrade codex 5.3 xhigh on such occasions? At least for pro subscribers. I had to rewrite pretty much everything today.
English
0
0
0
23
OpenAI
OpenAI@OpenAI·
5.4 sooner than you Think.
English
3.5K
2K
28.9K
5.1M
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
@GeminiApp Any chance of 3 flash lite api being not more expensive than 2.5 flash lite? Price creep limits usage of new models with tasks that don't need very smart model, but need a cheap model
English
0
0
0
146
Google Gemini
Google Gemini@GeminiApp·
Gemini 3 Flash is here ⚡️ Get the free, fast, unlimited version of Gemini with our biggest upgrade yet. See what’s new and how we’re using Gemini 3 Flash to tackle everyday tasks and get answers fast. (thread)
English
515
765
6.6K
7.2M
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
@Sauers_ Selfhosted runners on a 100 eur/mo hetzner instance give us way better experience
English
0
0
2
1.2K
Sauers
Sauers@Sauers_·
GitHub Actions gives everyone 13,440 CPU-hours for free per week ($30k annually in equivalent AWS compute). You are using these, right?
English
80
79
3K
271.2K
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
Gemini 3 Pro seems very smart so far but attention to details is a disaster.
English
1
0
0
26
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
I didn't update to MacOS 26 until 26.1. Upgraded today to find out that language switcher is still buggy. High quality software my ass...
English
0
0
0
24
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
@gdb Maybe the frontend team can use it to find out how to not annoy users by switching from pro model to auto after each message.
English
0
0
0
98
Greg Brockman
Greg Brockman@gdb·
Taste of what LLM-driven scientific discovery will be like: with 12 minutes of thinking, GPT-5 Pro suggested repurposing a known drug to treat an untreatable food allergy. Same exact result was found by an (at the time unpublished) peer-reviewed study. And models still improving.
Derya Unutmaz, MD@DeryaTR_

Here is the story of a remarkable, independent treatment suggestion by GPT-5 Pro: repurposing a known drug for a patient with food protein–induced enterocolitis syndrome (FPIES). First, how we came to test this. My close friend, physician-scientist Dr. Oral Alpan, treated the patient described in the published report (link in the thread) for refractory allergic skin disease with the biologic dupilumab, which is approved for that indication. The patient also had FPIES, a food allergy that in his case was triggered by wheat and caused hours of cramping and watery, sometimes bloody, diarrhea. There is no approved treatment for FPIES; patients are advised to avoid trigger foods and prepare for emergencies if accidental exposure occurs. For twenty years, even a small amount of wheat set off the same cascade a few hours later, followed by days of recovery. After starting dupilumab for his skin condition, the patient traveled to France and accidentally ate a baguette. To his surprise, nothing happened! It was his first uneventful wheat exposure in two decades! On return, Dr. Alpan hypothesized that dupilumab might be responsible and supervised an oral food challenge approaching 50 grams of wheat protein. Again, no reaction! When an insurance interruption forced a pause in medication, the old symptoms returned; restarting it restored tolerance. In the peer-reviewed paper published today, Dr. Alpan and his team describe seven additional patients, ages 2 to 58, who responded to dupilumab for their FPIES condition. While this is not definitive proof and represents an observational case series, the findings suggest that dupilumab could be a potential treatment for FPIES. Dr. Alpan intends to contact Regeneron, the manufacturer, to pursue clinical studies hopefully toward FDA approval for this disease. About two months ago, as he had just submitted the paper on this case, Dr. Alpan told me this story. Until today’s publication there were no reports in the literature of dupilumab potentially treating FPIES. Around that time GPT-5 Pro had just been released, and I thought this would be a perfect case to test whether AI could infer a treatment from the clinical case alone. I remember anxiously watching it think for about 12 minutes. When the output appeared, I almost fell out of my chair! Its top recommendation read: “ dupilumab (IL-4Rα blocker).” (see attached screenshot). Just like my friend had done, it was leveraging the skin condition to treat FPIES. A key insight could be that GPT-5 Pro had recognized this drug is not only effective for the skin condition but also has a “broader epithelial effect,” while this has not been proven in FPIES patients, I think it is a plausible mechanism in this disease. It also listed other options that were mostly symptomatic and flagged one with an unfavorable side-effect profile. There were several other suggestions, though those are most obvious ones and also suggested by other models too. No other model I tried identified dupilumab as a first choice; though GPT-5 Thinking suggested it as a second or third options. Many other treatments for currently untreatable conditions may already exist and can be uncovered by using GPT-5 Pro and other advanced AI models to repurpose drugs (among 20,000 FDA approved drugs!). Of course these will still need to be clinically tested but this is how AI can help catalyze a healthcare revolution! I am very excited about what comes next and future AI models will be even more powerful to make such discoveries! Important safety and regulatory information: Dupilumab is not approved by the U.S. FDA for the treatment of FPIES. In this study, dupilumab was used on-label to treat each participant’s comorbid, FDA-approved allergic indication; the favorable FPIES observations occurred in that clinical context and should be considered off-label with respect to FPIES. Patients should not initiate or change therapy without consulting their physicians. This press release is for informational purposes only and is not medical advice. Also, as evidence that this was done before today’s publication, I sent a GPT-5 Pro link to friends at @OpenAI about two weeks ago to give them a heads-up, but I waited until the paper went online today to reveal this story.

English
140
228
2K
561.9K
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
@OpenAI Can you please stop switching from pro to auto after each message? It's really annoying, I expect some respect to my choice of the model for $200.
English
0
0
0
2
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
@YouTube Even Gemini knows that it is a very common and frustrating experience. Can you stop pushing "1 language is enough for everyone" paradigm?
TheFullStackSnack tweet media
English
0
0
0
7
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
Every san non-appstore app on mac: I'm running, let's use this time to download updates MS Office on mac: it's my laptop now, go to setting and enable updater so I can run whenever I want, not whenever you want.
English
0
0
0
27
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
For example, I see a lazy rage baiting, but account seems kinda relevant to my interests at first glance. Now I have to make a decision right there whether to mute. I don't want to spend time analyzing post history, so the decision is semi-random. I'd rather give mute-strikes, and mute on 3.
English
0
0
0
6
Nikita Bier
Nikita Bier@nikitabier·
If an account is found to use AI to write automated replies, they should be:
English
1.5K
108
2.8K
539K
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
.env + .env.local > .env.sample
Nederlands
0
0
0
10
TheFullStackSnack
TheFullStackSnack@f_stack_snack·
diskonaut is no longer in nixos packages. Is there a good alternative (speed + tui)?
English
0
0
0
22
Santiago
Santiago@svpino·
I think @code needs to spend a bit more time improving their Tab model. That's their only disadvantage right now with Cursor. I find they are on par or better than Cursor everywhere else.
English
18
2
91
20.4K