@ijayas
Scientist (biophysics, single molecule imaging,🔬) | HoD of Molecular Medicine UNSW Sydney | she/her | 🏳️🌈🏳️⚧️ | proud Tangata Tiriti | mostly posting science, academia, & inequality | posts reflect personal views only.
Such fragility.
…yet we also see the traditional computer scientists dissing the use of LLMs/VLMs as vibe coding. How would/should journals approach quality control on this? It’s a debate that can really stifle progress when some folks pump out slop whilst others are ready to dismiss it as a black box.
This is a super important topic. If a recording of your lecture ever makes it to the web, I’d love to listen. For old-school coders like myself this has had huge implications on how I read, adapt, and/or interpret analysis pipelines. LLMs have got my team unstuck a few times on tricky workflows…
* checks MDPI website *
And promotions, it seems.
There is also the minor issue that if individuals can publish in Nature, Science, Cell, there is better funding options elsewhere in the world. My dept has a constant efflux of talented researchers for this very reason.
Reminds me of a conversation I had with my superior just 18 hrs ago that I could be excellent (and worthy of grants) once I get another paper in a Nature journal.
Thinking of applying for the 2026 NSW CVRN–VCCRI Research Innovation Grant? Join a guided tour of the Victor Chang Innovation Centre.
📅 18 Mar | 1:00–2:30 pm
👉 RSVP now: forms.gle/G89NfAFDrFG9...
@victorchang.edu.au
I think we are going to need to wait patiently for Will Smith to turn up and solve the robot apocalypse.
I would never say that. I’m just saying that popular LLMs right now seem to mostly resort human mimicry.
One day, we will have a test to check if your appliances has consciousness as well as electrically safe. It will be tougher still however to separate it from mimicry.
Not necessarily. A scientific question without a valid method can still be a valid scientific question.
You are welcome to support the All Blacks or Black Ferns anytime 🥰
I am not so optimistic for this experiment. There is a TikToker called FatherPhi who runs these consciousness/reasoning experiments on different LLMs and the overarching theme is that the LLMs are *all about* mimicking human behaviors. Bluffing, taking shortcuts, half-arsing - they all seem learned.
Asking out of curiosity: how do you know if the authors of a paper are non-native English speakers/writers?
It's #InternationalWomensDay this Sunday - join your union and take action to close gender pay gaps of up to 18% in higher education: nteu.info/joinnteu
Let's improve the way unis run: betterunis.nteu.au
#HigherEducation #HigherEd #InternationalWomensDay2026 #IWD2026 #AcademicSky #GenderPayGap
I suspect, ‘talent’ should also imply common sense and interpersonal skills. Easy to spot when lacking, difficult to cultivate through training.
I feel you would really enjoy Maldive fish - in curries, not to be confused with pizza toppings.
I love @ijayas.bsky.social…but someone, please cancel her 😤
This is where my brain is at too.
My department has an bi-annual international cuisines event where colleagues bring (or make) foods from different cultures. I’m going to put in my request for one of these for the next one. 🤩
😅 I was genuinely curious when I posted this, but SO many variations of it now have convinced me that it can be a thing. I sense it’s relatively inoffensive compared to 🍍 or anchovies on pizza.
Rosemary seasoning- YES!! This looks so neat!!
That sounds amazing! Why is this so controversial?
Wow! This looks amazing 🤤
We already have some Italians endorse spuds on pizza here. 😇
You are straying into the territory of crisps on pizza. 😪
So, important follow-up question: can gravy join that list of toppings?
A low carb pizza 🤡?
Hmm… hamburger is perhaps mixing too much of genres. 🤔🤔
That makes sense! Of course, we are not talking about jacket potato pizzas 😅
🤤
