(The post here reflects my own thoughts and may not be the thoughts of my employer, just putting that out there now to avoid any confusion)
The Shift
Over the last few months there’s been a shift. A movement from web sites and apps that do stuff (mostly useful, some utterly useless) to a move refined thinking on process and insight.
During the weekend I was look at the funding patterns of artificial intelligence startups. Handily KDNuggets (the place you look for anything on data mining and machine intelligence) had a piece on 50 of the “top” companies right now in AI.
The 50 to Watch
Company | Sector | Investment ($m) |
---|---|---|
InsideSales.com (Provo UT) | Ad Sales | 251.2 |
Persado (New York NY) | Ad Sales | 66 |
APPIER (Taipei Taiwan) | Ad Sales | 49 |
DrawBridge (San Mateo CA) | Ad Sales | 46 |
Zoox (Menlo Park CA) | Autotech | 290 |
Nauto Inc. (Palo Alto CA) | Autotech | 14.9 |
nuTonomy (Cambridge MA) | Autotech | 19.6 |
Dataminr (New York NY) | BI | 183.44 |
Trifacta (San Francisco CA) | BI | 76.3 |
Paxata (Redwood City CA) | BI | 60.99 |
DataRobot (Boston MA) | BI | 57.42 |
Context Relevant (Seattle WA) | BI | 44.3 |
Tamr (Cambridge MA) | BI | 41.2 |
CrowdFlower Inc. (San Francisco CA) | BI | 38 |
RapidMiner (Boston MA) | BI | 36 |
Logz.io (Tel Aviv Israel) | BI | 23.9 |
BloomReach (Mountain View CA) | Commerce | 97 |
Mobvoi Inc. (Beijing China) | Conversation AI | 71.62 |
x.ai (New York NY) | Conversation AI | 34.3 |
MindMeld (San Francisco CA) | Conversation AI | 15.4 |
Sentient Technologies (San Francisco CA) | Core AI | 135.78 |
Voyager Labs (Israel) | Core AI | 100 |
Ayasdi (Menlo Park CA) | Core AI | 106.35 |
Digital Reasoning (Franklin TN) | Core AI | 73.96 |
Vicarious (San Francisco CA) | Core AI | 72 |
Affectva (Waltham MA) | Core AI | 33.72 |
H20.ai (Mountain View CA) | Core AI | 33.6 |
CognitiveScale (Austin TX) | Core AI | 25 |
Numenta (Redwood City CA) | Core AI | 24 |
Cylance (Irvine CA) | Cyber Sec | 177 |
Darktrace (London UK) | Cyber Sec | 104.5 |
Sift science (San Francisco CA) | Cyber Sec | 53.6 |
Kensho (Cambridge MA) | Fintech | 67 |
Alphasense (San Francisco CA) | Fintech | 35 |
iCarbonX (Shenzhen China) | Healthcare | 199.48 |
Benevolent.AI (London UK) | Healthcare | 100 |
Babylon health (London UK) | Healthcare | 25 |
Zebra medical vision (Shefayim HaMerkaz Israel) | Healthcare | 20 |
Anki (San Francisco CA) | IOT | 157.5 |
Ubtech (Shenzhen China) | IOT | 120 |
Rokid (Hangzhou Zhejiang China) | IOT | 50 |
Sight Machine (San Francisco CA) | IOT | 44.15 |
Verdigris tech. (Moffett Field CA) | IOT | 16.1 |
Narrative science (Chicago IL) | Text Analysis | 29.4 |
Captricity (Oakland CA) | Vision | 51.9 |
Clarifai (New York NY) | Vision | 40 |
Orbital Insight Inc. (Mountain View CA) | Vision | 28.7 |
Chronocam (Paris France) | Vision | 18.35 |
Zymergen (Emeryville CA) | Other | 174.1 |
Blue river tech (Sunnyvale CA) | Other | 30.4 |
Key Summary
- Minimum Investment – $14.9m
- Maximum Investment – $290m
- Average Investment – $73.26m
- Number of companies listed – 50
The listed companies were “ones to watch”, that doesn’t take into account the other 10,000 or so that will be in stealth, not on anyone’s radar or just making sales and getting on with it.
For me one concern is the lower investment limit, $14.9m, I’ve not seen any NI company raise that amount of investment. And I’ve spent time thinking about why that could possibly be.
- All the startups are donkey’s. They’re just not worth that amount.
- All the founders are playing the Northern Ireland funding game, raised their little $1m and can’t raise as they’ve already lost 20-25% of the company.
- There’s no actual IP or product.
- There’s no customers.
- There’s no problem being solved.
That’s off the top of my head, if I really went all mind palace on it I’d probably come up with another ten reasons.
The Talent Pool
The much lauded reason for FDI companies setting up shop in Belfast and, occasionally, Derry.
“you have graduates – there’s a lot of talent in Belfast” From the BT, here.
Which I read as, “There’s plenty of cheap graduates looking for a job in Belfast, we can exploit that and reduce our bottom line.”
It’s time to seriously question this marketing message, yes there are some very talented graduates in Northern Ireland. Are they ready for the market where they are needed? Debatable. Do they fill the gap of what’s really missing, no they don’t.
It still skirts around the issue for any startup, a complete lack of good CTO talent. What I’m seeing more and more of are companies setting up, getting that free government money (startup DLA, if you will) and handing out vanity titles like there’s no tomorrow. I’ve written and spoken about this many times before, if you want to read it again then have a look at this.
Good CTO’s in NI are hard to find, plain and simple. The reason for this is simple too, they’re pretty much in great jobs with large employers with a deal too good to lose and don’t think it was a fluke, the large companies engineer it that way, they obviously don’t want to lose good talent when they see it.
Jumping to a startup with a very questionable runway is a huge risk. Look at yourself in the mirror and ask yourself, “Am I worth the risk to my employees, my C levels and most importantly my customers?”.
If you flinch or can’t do it then you obviously need a session with Wendy Rhodes.
NI Needs a BIG WIN
If you think you’re on the starting wave of AI technology then you’re already five years too late. The same mistake was made with BigData opportunities. What I personally believe is required right now is for someone to bring a product along that is so unique and solves a problem better than anyone else that the rest of the world can’t do anything but look.
This thing also needs to IPO big time and make the founders and early stage investors so rich that people look at Northern Ireland as the place. The time is now to stop kidding ourselves and thinking we’re at the start of a wave, you’re already behind. Still thinking that social media data is going to make you (and others) rich, I doubt it, that edge is long gone.
There’s little point building tools, it’s hard to create revenue with programming tools and API’s, solve a problem better than anyone else so it can’t be ignored. The tools to do AI and Machine Learning are plentiful, whether it be TensorFlow, Weka or what have you. Search hard enough here and you’ll find posts on those technologies. At the end of the day the programming side isn’t that difficult when you have good coders who understand the logic.
I firmly believe it can be done, I just think the thinking needs to change, stop listening to salary paid government PR (use them, fine, but weigh up what’s being said) and focus on idea, IP and customer.
- Kick ass product
- Kick ass team
- More than $7m in investment
- An edge that no one can ignore.
- Main focus to remain in NI and IPO.
Your focus needs to be three standard deviations to the mean, that’s where the risk and the potential rewards are.
And keep this in mind, AI is not about replacing jobs, it’s about focusing on the job creation and creating new jobs that currently don’t exist. It’s an exciting time to be here but NI but you have some serious catch up to do.
Beltech 2017
I’m on the panel at Beltech 2017, “Public Debate: The Impact of AI on our World” at 6pm though I’ll be there most of the day on behalf of Mastodon C. So feel free to catch up with me there.