[Tweeters] (Correct Link)-Well Tweeters, what do think of this?: “World's first-ever smart binoculars can identify 9,000 birds thanks to built-in AI | Digital Camera World”

[Kevin Lucas]

"Merlin says" isn't sufficient justification. I do think it's valuable to
put in my checklists what Merlin identified a bird as, and whether it's
based on an audio recording or image analysis, and whether I agree. In
recent bird watching in Texas, using the Merlin app on my phone I recorded
Ladder-backed Woodpeckers repeatedly. I was watching the birds calling.
Every single time Merlin mis-identified Ladder-backs as being Downy
Woodpeckers. I could tell the difference by ear, and easily by sight. I put
that in each of my checklists. I think including that information in my
checklists is valuable. (I also corresponded with a local eBird reviewer
there about it.) Similarly, when I'm reading others' checklist reports of
rare birds, even with written descriptions that would fit, I consider the
observer's credibility since I've learned here that what is reported can be
both mistaken and falsified, sometimes with photos used of a different bird
or from a different time or at a different location, even by experts.
Merlin brings to my attention birds I'm hearing that I'm not noticing. And
when Merlin is misidentifying birds, as with nuthatches and woodpeckers in
Texas, I'm at first frustrated, then interested in why it's making the
mistake -- does it have the same tempo, overall pitch, melody,...? Is the
frequency response of my phone's microphone throwing it a curve? It gets me
thinking and listening differently. I don't just tick a box then move on. I
continue to listen more keenly. Is Merlin's bird actually there too? Am I
mistaken in my identification? Is it hearing something that I'm not? I
think I'd do the same with these AI binocs.
Years back on a Toppenish winter raptor field trip two local experts were
talking about a hawk, identifying it differently than I was. They were
looking through a Pentax and a Swarovski scope. I kept disagreeing,
describing what I saw and my identification. Finally, annoyed, one of them
came over and looked through my Zeiss scope. After fussing about for a bit
to be sure it was the same hawk on the same power pole, he realized that
the image in my scope was much better, and that their identification was
wrong. I couldn't imagine why they were calling it what they were. They
thought that I was wrong, maybe based on my lack of experience or their
very high self-confidence. They were the local experts. I simply had a
different, and better, view.
Sometimes only one or two of my photos in a series of hundreds reveals a
key field mark, or shows true colors, with others often misleading. I truly
enjoy going through them to see what seems whacked. I spend loads of time
watching single birds outside, observing how different aspects of their
appearance change. That's the best part of birding for me.
Accepting an identification as correct or true without question is
inappropriate whether from Merlin, Swarovski, beginners, or experienced and
trusted ornithologists or competitive birders. AI might catch a fleeting
bit that nails it, missed by even the fastest, best birder. IBM's Deep Blue
beat Gary Kasparov, a chess master, back in 1997.
I often note in my checklists when I think a bird has been misidentified by
Merlin or by people. Cornell's presentations about Merlin are clear that
misidentification is part of their program. Good birders recognize and
accept that about themselves too.
Besides giving credit where it's truly due, one of the reasons I want to
know who the person is who found a rarity is because sometimes I ask them
what made them realize it was a rarity -- not just the field marks, but HOW
they were birdwatching, or what they were focusing on. How could I change
my approach so I'm more likely to see something unusual or new?
I don't see these Swarovski binocs effecting that just yet, but maybe
someday. Before the rarity came into view, what other birds were seen? They
could have the GPS time/date/location embedded,.... That could be hacked to
falsify I suppose, like swapping out a photo of a Cooper's for a Sharpie in
the newsletter, or putting today's bird photo into a checklist created a
week earlier, but we might learn some new ways of looking for and
identifying birds. AI might also help catch frauds and cheats.
Catherine Hamilton, an artist drawing stints and peeps, noticed a
difference in their gape notch that I think hadn't been recognized before.
I thought that was awesome.
Artificial Intelligence seems likely to similarly come up with loads more
for us to recognize and realize and ways to see differently.
I can't run out and buy a pair of these AI binocs, but I'd love to play
with them to see what I see.
Even better, I'd love to have them as a spare on my shoulder to hand to
folks who ask me, "What are you looking at?" Sharing my digiscoping / phone
scoping display view elicits wondrous responses.
Seeing those responses is right up there with my watching a birds'
appearance change so amazingly as I watch. Oh, what we could share.

Here's a link to an article about Catherine Hamilton's stints and peeps
gape notch observation:
https://blog.aba.org/2017/11/open-mic-a-new-field-mark-for-differentiating-stints-and-peeps.html

Good Birding,
https://www.aba.org/aba-code-of-birding-ethics/
Kevin Lucas
Naturally Learning
more and more,
mostly in Yakima County, WA
*Qui tacet consentire videtur*
I apologize for inivitable errors.
-K
Yep. The last one was intintional.
The more eyes the better.


On Thu, Jan 18, 2024 at 10:59 AM HAL MICHAEL <ucd880 at comcast.net> wrote:


> One of my big concerns about the growth in AI and bird ID apps is that all

> they do is tell you what an object might be.  One doesn't need to learn

> behavior, field marks, jazz, and so on.  Just point and shoot.  One of the

> things drilled into me with ID (of anything) is "Why do you think it is

> X?".  For me, "Merlin says" is not acceptable.

>

>

> Hal Michael

> Board of Directors,Ecologists Without Borders (http://ecowb.org/)

> Olympia WA

> 360-459-4005

> 360-791-7702 (C)

> ucd880 at comcast.net

>

> > On 01/18/2024 8:25 AM PST Dennis Paulson <dennispaulson at comcast.net>

> wrote:

> >

> >

> > Hmm, the loss of all neurons associated with learning to identify birds?

> The death of the bird-book industry? No more bird ID classes? Roger Tory

> Peterson turning over in his grave? But I presume the price of those things

> won’t have much of that happening very soon.

> >

> > Dennis Paulson

> > Seattle

> >

> > > On Jan 18, 2024, at 7:45 AM, Dan Reiff <dan.owl.reiff at gmail.com>

> wrote:

> > >

> > > Hello Tweeters,

> > > Any other year, I would check the date to see if was April 1st!

> > > But, another way AI will continue to surprise us all in the many

> unexpected ways it will affect our options and lives. And faster than I

> would have imagined.

> > > What’s your reaction to reading about these new high-end binoculars?

> > > Dan Reiff, PhD

> > > 

> > > The correct link:

> > >

> > >

> https://www.digitalcameraworld.com/news/worlds-first-ever-smart-binoculars-can-identify-up-to-9000-birds-thanks-built-in-ai

> > >

> > > Sent from my iPhone

> > > _______________________________________________

> > > Tweeters mailing list

> > > Tweeters at u.washington.edu

> > > http://mailman11.u.washington.edu/mailman/listinfo/tweeters

> >

> > _______________________________________________

> > Tweeters mailing list

> > Tweeters at u.washington.edu

> > http://mailman11.u.washington.edu/mailman/listinfo/tweeters

> _______________________________________________

> Tweeters mailing list

> Tweeters at u.washington.edu

> http://mailman11.u.washington.edu/mailman/listinfo/tweeters

>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman11.u.washington.edu/pipermail/tweeters/attachments/20240118/e2b86350/attachment.html>